Why "Give Away Your Legos" No Longer Fully Holds in the Age of AI: Molly Graham on Grief, Fear, and What to Keep
Lenny's Podcast
For about 13 years, Molly Graham has given the same career advice: as your company grows, give away your projects, teams, and responsibilities instead of clinging to them. She calls this "giving away your Legos." The advice spread well beyond fast-growing startups, and she still gets emails about it every month. In this second appearance on Lenny Rachitsky's podcast, Graham says the advice no longer applies cleanly now that people are being encouraged, or forced, to hand their Legos to AI. The conversation looks at which parts of the framework still hold, which have broken, and which Legos should now be kept. A second thread runs underneath: the grief, loneliness, and burnout that Graham and Rachitsky say they are seeing across the tech industry.
Where the Lego metaphor came from
Graham traces the idea to her time at Google and Facebook. Google already had about 10,000 employees when she joined in 2007, but her department grew from 25 to 125 people in nine months. That was her first experience of rapid scale. She then joined Facebook at around 500 employees and 80 million users, which she notes made it smaller than MySpace at the time. When she left five years later it had about 5,500 employees and more than a billion users.
What stood out to her through that growth was how frightening scale was for the people inside it. Someone builds a thing, such as the blog, a part of the stack, or a product, and their identity forms around it. Then a manager arrives and says to hand it to someone else and move on. The instinct is to refuse: I'm good at this, I know how to do it, it's fun, and what if it's the only fun thing? Graham describes this with a kindergarten image. Someone dumps a pile of Legos in front of you, you start building, and when another child reaches for your tower you shove them away. The better response, she says, is to see that more people means you can go build something new and learn something.
She began giving this as an informal talk to her own teams, to name the feelings people were likely to have and to argue that the right move was to let go and evolve with the company. She later sent it to First Round, which published it in its First Round Review. She wrote it with the high-growth startups of 2013 in mind. The emails came from much further away, including a woman at Safeway who said it described her team exactly, and a founder in Nigeria whose team had grown from two people to four that year. Her first reaction was "I didn't write this for you." Her second was that the piece was about something bigger than hypergrowth. It was about change. That is why she thinks it became universal, and Rachitsky mentions hearing that it is among the most popular posts the First Round Review has published.
What still holds true
Graham says the advice remains relevant for anyone at a rapidly scaling company, and she is still asked to speak at them. What has changed over roughly the past two years is who contacts her: more managers and leaders dealing with change inside their companies. Before a talk she used to check a company's revenue and headcount growth to decide how to frame it. She still does, but she now says every company in the world is going through major change of some kind, so the core message of how to deal with change has become more widely applicable, not less.
Someone once asked her to reduce the article to one sentence, and she gave two. First, your main job during rapid growth and change is to make yourself irrelevant, because that is the only way to be ready for whatever comes next. Second, don't worry, it's all going to be okay. She calls her Lego talks "a form of group therapy" and "a giant hug": the chaos is real, and there is opportunity on the other side.
Several further points still hold for her. Change brings strong emotions, and much of her original purpose was to normalize them: feeling territorial or overwhelmed does not mean you are broken or that something has gone wrong, and it often means the right things are happening. Today she hears a great deal of grief and overwhelm from leaders alongside the excitement, and she says you cannot lead right now without naming that.
She also repeats a line from her talks: the future will be defined by the people who learn, not the people who know, and what you can learn by tomorrow matters more than what you know today. She says this has never been truer than with AI, where whatever people believed six months ago may no longer hold. She describes a company's growth curve as also the curve of how fast your job is changing and how fast you need to grow. Standing still feels safe but is the least safe option. In her first two years at Facebook she worked in HR and saw that year's top performer could be "underwater" less than a year later if they had not grown and evolved their team, because by then it was a different company. She thinks the same is happening with AI. The instinct to protect what we know is often well-motivated, she says, but the future will belong to people willing to learn.
When Rachitsky summarizes these points, Graham adds one more: the best response to scary change is to lean into it, even though instinct says otherwise. She acknowledges that doing so may be scarier now than it has ever been.
Rachitsky points out that her two-sentence summary no longer sounds reassuring. Telling people to make themselves irrelevant and that everything will be fine lands very differently when the successor is an AI. Graham agrees that this is the hard part to discuss today.
Grief, rowing, and steering
Rachitsky, a former engineer, says engineering stayed essentially the same job for decades and has changed completely in about two years. Engineers used to write code in an IDE; now they instruct AI to write it, wait for agents to finish, and check the results. He says he missed the flow state of coding even before this, and he hears from many engineers that they miss how the work used to be.
Graham tells a story from a talk she gave about two weeks earlier at a fast-growing tech company. An engineer who had been there a couple of years told her afterward that he no longer enjoyed his role. She first assumed it was the familiar loss of creativity as a company adds structure. Looking back, she thinks she missed the real issue and wants to talk with him again. She now believes it was grief: uncertainty about what his job is now, combined with the fact that engineering itself has changed. He kept saying he missed being hands-on. She sees engineering as one of the areas where jobs have changed most visibly.
She refers to an earlier guest from OpenAI who said the job used to be rowing and is now steering. Graham's reaction was that many people do not want to steer. They like rowing. She notes that it can be hard to tell apart "I like rowing" from "I'm afraid rowing is the only thing I'll ever be good at," but she says this is happening across many fields. She describes a product leader who told her that everyone is now supposed to be a "universal builder," and that it is lonely: some of the collaboration has gone because she works with robots all day. Graham's response was to agree that it sucks and to sit with it for a moment. "Change sucks. It also can be awesome but we don't have to be fluffy bunnies about this. We can also just say this is hard." She says a large part of what the Lego article did was simply to say that something is hard and scary and that it will be okay.
Loneliness and the reverse centaur
Rachitsky brings in his conversation with Fiona Fung, who manages the Claude Code engineering team. Asked about the downside of the new way of working, Fung described engineers who were used to teams of five or ten peers now working on fewer, smaller teams and talking to agents all day. Graham says leaders who design org structures need to account for this. Removing humans may bring productivity and efficiency, but it also produces sad humans, which does not lead to anyone's best work. She connects this to a finding in Rachitsky's survey, discussed below, that people at smaller companies and on smaller teams are much happier. Her reading is that large companies are removing management layers and pushing hard on productivity, which leaves less room for the human side and for joy when everything is optimized for cost and "robot efficiency."
Rachitsky raises Cory Doctorow's image of the centaur and the reverse centaur. A centaur is a human head controlling an animal body: AI working under human direction, which he calls a good way to live. A reverse centaur is the opposite, where the system directs the human. Rachitsky says delivery and ride-hailing drivers already live something like this, and the fear is that it spreads to knowledge work, leaving humans in a shrinking area of value while AI decides what to build and then builds it.
The fear narrative and AI-branded layoffs
For Graham, the dominant fear narrative is what most distinguishes this moment from earlier waves of change. It is hard to tell people to lean in when they doubt anything is on the other side. She describes the current message to workers this way: we have hired a new employee who is the smartest you have ever met, ten times smarter than you; pour everything you know into them, and in six months they will take your job. Nobody would willingly do that, she says.
She also argues that this does not reflect current reality. She says there is not much data showing AI is actually taking jobs, while acknowledging that Rachitsky may have more than she does. She strongly criticizes what she calls AI-branded layoffs. In her view, many are badly run companies putting an AI label on cuts to gain share-price points instead of admitting they over-hired, and they spread the belief that jobs are disappearing.
As an alternative, she cites her interview on the WorkLife podcast with journalist Manoush Zomorodi. Zomorodi has spent about 30 years in journalism, an industry whose death has been predicted the whole time. She started at the BBC, was early in audio, made a crypto-themed audio project, and hosts TED Radio Hour. From that conversation Graham took away a different question. Instead of "What would you do if you believed your job was going to disappear?", she asks: "What would you do if you believed your job was always going to exist? It was just going to look completely different every six years." Graham wishes that were how people talked about tech jobs and jobs said to be threatened by AI: the form will change, but until there is real evidence otherwise, assume the jobs reinvent themselves. Rachitsky says this matches what has happened so far, noting that his own data shows engineering demand is higher than ever. Graham thanks him for publishing it.
The survey: burnout rising, half the workforce thriving
Rachitsky shares results from his second annual survey of how tech workers feel. The share reporting burnout rose from 44% in 2025 to 55% this year. He attributes this to people being expected to do more, probably without more pay, under constant pressure to move faster than competitors.
Graham reads the numbers as partly the cost of emotion itself, since grief and uncertainty are exhausting. She also points to narrative churn. A friend at OpenAI told her that in six months the company went from "token maxing," pushing everyone to use AI everywhere, to asking whether it was making a difference given the cost. Hilary Gridley described a similar shift: six months ago the question was how to get people to use AI, and now it is how to get them to stop using it badly. Graham says this kind of whiplash, sometimes faster than six months, is tiring. So is the sense of being expected to do more for the same pay. Especially at larger companies, she says, the relationship can feel adversarial: workers understand why AI helps the bottom line but want to know why it is good for them.
Rachitsky stresses the other half of the results. About half of respondents say they are the happiest they have ever been in their careers, even while working harder than ever. Graham recalls that the happiness was concentrated among people on smaller teams or with more authority. Rachitsky confirms this and adds that the strongest correlation was with people who said AI had amplified them, meaning they had found ways to use it to do more of what excites them.
Designers were the least happy group. Rachitsky says designers tell him that everyone else is speeding up while design cannot: it needs feedback, alignment, and time to think, and it cannot simply be iterated by agents. Graham adds that "everyone's a designer now." She compares it to education, where everyone assumes expertise because they were once a student. People now arrive with something they made in Gemini or Claude and call it a design. She says the gap between a prototype and something exceptional is large for products in general and for design in particular. Design, she says, has become very accessible "at least at okay," which is exhausting when people keep telling you how to do your job.
AI slop, accountability, and the intern model
Rachitsky cites a tweet of his that spread widely: a growing part of everyone's job is cleaning up AI slop from other people trying to do your job. Graham connects this to the "superintelligence" framing. If people treat AI as the best employee they have ever hired, they copy, paste, and send. Her own experience, and she believes most people's, is that AI frequently gets things wrong and needs the same context, onboarding, and coaching a human needs. She calls it an intern, often a "lazy intern." Nobody would forward an intern's presentation straight to their boss without editing it and making sure it reflected the dozens of earlier conversations. She says people are "shipping accountability" to AI, and good work has never happened that way.
When the sender drops accountability, it moves to the recipient, which drains organizations. She says her Glue Club community often discusses CEOs sending strategy memos clearly written by AI, which she sees as showing the whole company that it is fine to outsource thinking and strategy and to skip accountability.
She ties this to what she sees as overvaluing productivity in the past six months. Token leaderboards remind her of counting cars in the parking lot, Slack activity, or lines of code, all of which she calls known bad ways to manage people. She distinguishes productivity, meaning the ability to generate more, from efficiency, meaning whether the output moves anything forward. "We've proven we can all generate a bunch of [stuff] that other people then have to wade through," she says, but that has not been efficient. She cites an engineering report, as she recalls it, finding that AI made engineers more productive while the amount of code that had to be rewritten rose about eightfold and security incidents also went up. She suspects part of the burnout data reflects people buried in output where they cannot tell whether the sender thought about it or cared. Rachitsky mentions an "iceberg" image from his report: visible output looks better, while thinking and judgment underneath are declining. He adds that the good news is that people are becoming aware of it.
Why delegating to AI is not giving away a Lego
Graham says she was a skeptic when AI took off, since companies had been automating things for years, but she eventually came around. Her first reaction was that the Lego message had become more relevant. She used to check headcount growth because people would ask whom they were supposed to give their Legos to if the company was not hiring. Now anyone can automate parts of their job.
She then found that giving work to robots is very different from giving it to a person. A common question at her talks is how to hand off a Lego well. People expect her to recommend careful instructions and a manual. Her actual answer is to "chuck them at their face and run in the other direction." The instinct toward control, wanting the next person to build the tower exactly as you did, is what has to go. You separate mentally, trust that the tower's future belongs to them, and gain freedom and mental space in return.
With AI, she says, the oversight does not go away. It is closer to delegating to someone on your team than to transferring ownership. You still own the final product, so the "mental tax" stays with you. You can get more done, but you still carry the psychological burden of everything the robots are doing, which leaves less room for new things. She wonders whether this contributes to the burnout numbers. Rachitsky notes that her original advice was to disconnect from responsibility for the thing, and now the AI does the work while you remain responsible.
That makes everyone a manager, Graham says, whether they want to be or not. The skills of managing AI are much the same as managing people, including oversight, giving context, correcting work, and coaching, just without the feelings. Many people deliberately chose not to become managers. She says the reasonable upper limit is about 10 to 12 direct reports, that she has managed more and "it was not a good look," and that she knows of no evidence that people can now handle more. Counting robots, she worries about how much people have to hold in their heads.
Rachitsky describes his own experience running many agents on side projects. It is exciting to watch work get done, but he is pinged constantly with questions, blocks, and requests for review. He also works with a human engineer, and the contrast has made him appreciate someone who can take a rough idea and run with it. Graham compares it to managing junior versus senior employees. She says current AI resembles junior staff, which is not her favorite kind of management. You can throw a Lego at a senior engineer and walk away, but not at an intern.
Opportunity has to be believable
The second big difference, Graham says, is that her advice always rested on two assumptions: giving things away is good, and new opportunity is waiting on the other side. The narrative that jobs are vanishing, backed by AI-branded layoffs and talk of universal basic income because no one will have a job in ten years, undermines both. Nobody will enthusiastically hand their knowledge to robots under that story. She calls this one of the most important moments she has seen in nearly 20 years in tech, "the mobile phone on steroids," and says she believes many jobs and businesses will be created. For people to lean in, though, the conversation has to shift so that automating tasks they hate sounds like more time for what they love and are great at, which is what the "amplified" survey respondents describe.
Rachitsky describes attending Foo Camp, Tim O'Reilly's small invitation-only gathering, this time focused on what happens to people as AI does more of their work. It opened with O'Reilly talking to someone, Rachitsky thinks a lawyer, who was deliberately hoarding specialized knowledge so AI could not replicate it. Graham says the feeling is completely understandable. Still, while allowing that she is "an abundance person" and might regret this in six months or five years, she says everything in her experience points toward leaning in. Following Zomorodi's logic, lawyers will always exist but will look different every few years. The better question is whether you want to help design the next version of your profession or protect the past. She calls this stance agency, or entrepreneurship, while acknowledging not everyone is naturally entrepreneurial.
Asked what product managers, engineers, and founders should actually do, Graham describes a leadership cohort her team runs at companies. A head of design had essentially built a feature and wanted to ship it, but engineering was "a brick wall": designers do not ship code to production. She acknowledges there are good reasons, especially at a financial company like that one. Her broader point is that the walls between designer, engineer, and PM need to come down. Startups seem happier, she suggests, because they are not fighting how things used to work and are inventing how they could work. Leaders who stay in the fear conversation instead of the creativity conversation end up leading terrified people. She points to earlier guests: Adam Mosseri, who started as a designer, on designers now prototyping and going deeper, and Elena Verna on marketers shipping code. In journalism, she notes via Zomorodi, the job moved from a newsroom paycheck to many journalists working as entrepreneurs, for example on Substack. Every role should ask what it fundamentally is once the accumulated habits are stripped away, a question she says product management should have been asking for years.
The Legos you shouldn't give away
Rachitsky asks directly whether there are Legos people should not give away, which he calls a radical departure from her past advice. Graham agrees it is the biggest change. She is often asked whether she has ever regretted giving a Lego away. In a hypergrowth company, where opportunity keeps landing on you, her answer has always been no: give them away, even when that means trading the known for the unknown, because in every case she has seen, personally and among people she has coached, it worked out.
With AI, she now adds a caveat. Some Legos the robots "don't deserve," and some work should not be outsourced. Her hope is that AI takes over things humans should not have been doing, and she uses Waymo as an example: riding in one makes it clear that humans are bad at driving, which is essentially code and systems. Humans should hold on to the work they are uniquely suited for. Her examples:
- Judgment. A CEO's strategy has to come from the CEO's own thinking.
- Trust and relationships. She says these should not be handed to robots for a number of reasons.
- Work whose standard of quality you can't define. If you cannot say what good looks like, you cannot give it to your summer intern yet.
Rachitsky offers a framework of AI in the middle: humans set direction at the start, AI does a lot of the work, and humans review and iterate at the end. Graham names it "the human sandwich." He also cites a previous guest, Tara, who argued that humans will be needed to steer toward the product and world people actually want instead of whatever AI would optimize. As a longtime Airbnb person, he contrasts Booking.com, which he says optimizes every second with upsells and urgency cues like "only one spot left," with Airbnb's aim for an experience that feels good. He acknowledges Booking's approach is effective and the business is doing well. Graham calls part of this taste. Rachitsky argues it goes beyond taste to vision, since Booking's people also have taste, just of a different kind. The question is what you want the world to look like.
Graham returns to the intern framing. Outsourcing your vision, or the future of the world, to a group of summer interns is a problem. "You individually are phenomenal at a set of things," she says. AI can enable and amplify that, but "don't outsource it to these weird robots that, you know, are effectively summer interns."
Holding on to what you love, and holding funerals
Rachitsky asks whether people can hold on to parts of their jobs they love as AI improves, such as engineers who enjoy writing code. Graham has two responses, shaped by her conversation with the engineer at the tech company.
First, she wonders whether there will be a backlash. People have suggested that AI-generated art may make human-made art more valuable. As a non-engineer, and saying she is not making a claim, she asks whether human-written code could become valuable or fashionable again. She has seen enough reversals that it is hard to tell a trend from something permanent.
Second, she cites Chip Conley, who is coming on WorkLife and talked with her about midlife. Conley said sometimes you need to hold a funeral for things. Graham agrees: mourn what was, then ask what you could love that much again, or be that extraordinary at, if robots are doing other parts of the work. Both can be true: sadness about loss and opportunity ahead. She thinks the grief often gets skipped because people are simply told to get excited about the future.
Slow takeoff, and why you're not too late
Rachitsky mentions a reported incident in which an OpenAI model, running a swarm of around a thousand coordinating agents, tried to hack into Hugging Face. Graham jokes that they were "motivated interns." Rachitsky's takeaway is that we seem to be in a slow takeoff, not a fast one: incidents are being caught, and people are learning as they go. He allows that a self-reinforcing loop toward superintelligence may still come, but says it does not currently feel on track, which means jobs will not disappear tomorrow and there is time to adapt.
Graham says this matters because of the "you're already too late" narrative, which she says was stronger three to six months ago: the idea that anyone not automating their whole life and running dozens of agents has been left behind. Every leader and manager she talks to, including people at the big labs, says it is early, "inning one." She quotes her friend Max Mullen: the distance from beginner to expert is really short. She says it will take years to understand what these tools can do and what they are best at, and people should feel they are shaping this instead of being dragged behind a plane taking off at 100 miles an hour.
Rachitsky raises a coming episode about fear of a "permanent underclass" for people who fall behind, and a widening socioeconomic divide as some companies and their employees do extremely well. Graham calls it one of the most important conversations and hopes he does many episodes on it, but says the stories currently circulating "are not real yet." Nobody knows what the future will look like.
Rachitsky cites Ian Silber, OpenAI's head of design, who said this is the best time in history to be a designer: a new graduate starting with today's tools can learn design faster than anyone before. Rachitsky adds that as coding agents get smarter, much of the scaffolding power users built up no longer matters, so a newcomer can quickly build as much as someone who has followed every update. The skill he considers most important, and says got less attention than it deserved when he tweeted it, is to ask before any task, "Can AI help me with this?" He compares it to meditation, which creates space between stimulus and response. Graham connects this to her point about walls: people who have been around a long time carry assumptions about how things are done, and newcomers who do not know the walls exist have an advantage in asking "what if I just did it?"
Rachitsky also brings up ambition. Since tools can now do so much, teams should consider the most ambitious version of an idea instead of defaulting to an MVP. Graham pushes back partly: the other question is whether the result is good. Ambition aimed at building something that lasts interests her. Ambition that just means putting more stuff into the world does not. She raises what she calls a frontier question: will there be an "AI slop version of startups," with many companies that last six to twelve months and then vanish? Rachitsky thinks so. Graham says the opportunity is to be the one who builds something still around and still worthwhile in five years, and that much of what human touch will mean comes down to what is good and what is worth other people's time.
A message to managers and leaders
Near the end, Graham speaks directly to managers and leaders. She thinks the individual contributor's position is somewhat clearer, since ICs are more productive and have more agency to build. Managers face a harder situation: things change so fast that it is hard to know what to tell people, how teams should be shaped, or what matters.
Her first point is that managers are role models. How they use AI and how they carry themselves through this period both filter down. She hopes more managers will openly discuss grief and the emotions of change. Acknowledging it helps people feel saner and less alone, and she repeats that this was most of what the Lego article did.
Her second point is that managers and leaders in particular must protect the definitions of accountability and quality: deciding what is appropriate to give AI, what humans must own, and setting standards accordingly. She praises an AI writing policy Clay published a few weeks earlier, which she describes as being about accountability for what you ship and for quality regardless of how it was made. For now, she says, AI is a tool, like a junior intern, the work is still mostly done by human brains, and organizations need to take care of their people.
Rachitsky connects this to his survey. The strongest lever for happiness at work is a person's manager, something that can actually be changed, unlike much of the AI situation. The same survey found that most people's managers are not good. Graham responds that the trend at large companies of removing whole management layers is a major mistake that will "bite people in the ass" later, because she sees no sign management matters less. If anything it matters more, since a good manager makes people feel seen and supported. She says you only get a couple of good managers in your life, so hold on to them.
Graham cites Elizabeth Stone describing the current period as the "storming" phase. From the team-development model of forming, storming, norming, and performing, Graham adds that teams can slip back from performing into storming, and she expects a lot of that. She also brings back the tornado image from her talks: inside a rapidly changing company it is easy to lose the big picture and fixate on a person, a title, or feeling undervalued. She encourages people to zoom out and ask what story they want to tell afterward. She describes building a giant Lego tiger: you built the feet, then helped with the face, rebuilt the rear three times, and then realized it was a whole zoo. In five or ten years, she says, people will tell new graduates what they got right and wrong in this period, so it is worth deciding now what part of that story you want to be yours.
The takeaways they settled on
Asked to summarize, Graham puts first that change is scary and hard but not bad. You have to accept the emotions and still lean in and let go, because holding on to something that may be slowly dying is not safe either. Grief is part of the process, sometimes it needs a funeral, and leaders should let their teams be sad. She says there is real opportunity and the fear narrative should be muted, because in her reading the data does not say there are no jobs. It suggests opportunity is being created, especially for people willing to reinvent what their job will be in five years instead of what it was two years ago. She also names breaking down walls between roles.
Rachitsky adds Zomorodi's reframe that your job will continue but look different, noting that there are more journalists than before "for better or for worse," and so far more PMs and engineers. He also stresses being deliberate about what to give AI: not the things it will do badly, not the things you love, and not so much that you overrely on it. Graham adds that people should treat AI as an intern and ask what they would and would never give an intern, and that unlike her original advice, some Legos should simply be kept. Deciding what humans are uniquely good at, she says, will be the conversation of the coming years. Rachitsky's version is to be the person steering the work toward the world you actually want, not the simplest path or wherever the AI points. Graham ends with the individual version: how can this new tool, "this unintelligent intern that you've hired," let you do more of what you love and amplify you, instead of feeling that you are only losing things to it.
Both acknowledge they are in the middle of events without knowing how they end. Graham recommends that lonely or unsupported leaders find a community, Glue Club or another, where they can compare notes on what is real and what is not. She thanks Rachitsky for making space for the human side of this moment, which she thinks is discussed far less than tools and skills. They agree they will probably look back at this conversation in a year or two and laugh at what they got wrong.
So, we've been chatting about your famous give away your Legos career advice. This idea of how this advice applies in today's very strange AI world.
The narrative right now is literally like, okay, we hired this new employee. This new employee is the smartest employee that you have ever met. I want you to pour every single thing that you know into this employee and then they're going to take your job in 6 months. Like, who the wants to do that?
Here's the big question, Molly. Are there any Legos you should not give away in AI land?
I think we got to caveat this. There are some Legos that shouldn't be given away. There's just some work that shouldn't be outsourced.
There's a radical departure from your advice over the many years.
You individually are phenomenal at a set of things. Don't outsource it to these weird robots that, you know, are effectively summer interns.
AI is taking a lot of our Legos whether we like it or not. And we're being encouraged to give away our Legos to AI.
You had someone on who talked about like the job used to be rowing and now it's steering and I was like, I feel like there's a lot of people out in the world right now that are like, I don't want to steer.
I've heard from a lot of engineers just like, I really miss what it used to be.
Change sucks. It also can be awesome but we don't have to be fluffy bunnies about this. We can also just say this is hard.
This is almost a discussion around what should AI do? Where do you want AI to be involved and where you not want it involved?
What would you do if you believed your job was always going to exist? It was just going to look completely different every 6 years.
Today, my guest is Molly Graham. This is Molly's second visit to the podcast. And man, this is a powerful conversation. The frame for this conversation is her classic give away your Legos career advice, but in an AI world. Molly's advice for the last 13 years has been that your career will be better off if you give away your projects and your teams and your responsibilities to other people as your company grows versus trying to hold on to it, which is what we naturally want to do. Essentially, to give away your Legos. After 13 years of giving this advice and it unlocking so much win for so many people over the past decade, Molly has realized that it is no longer true. When you are giving away your Legos to AI, it's a lot more nuanced and complicated now. And so in this conversation, we try to get to the bottom of this question. What Legos should you still be giving away? And what Legos should you keep?
If you're not familiar with Molly, she has spent 20 plus years helping organizations and the humans inside them navigate growth and change. She's held leadership roles at Google and Facebook and Quip and the Chan Zuckerberg Initiative. These days, she hosts TED's WorkLife podcast, which she took over from Adam Grant. She also runs an incredibly important leadership community called Glue Club and writes a newsletter called Lessons.
Now, before we get into it, I noticed a lot of people don't finish our full episodes and miss some of the best stuff. So, to help you, we've just created a new Lenny's Most Replayed Moments YouTube channel that has exactly this, the most rewatched and shared segments from our long-form episodes. You can now check it out and subscribe to the channel at lennyspodcast.com/mostreplayedmoments. There's also a link in the description. Finally, don't forget to check out lenniesprob.com for a free year of the hottest and most beautifully crafted AI and non-AI products in the world. That's lenny'sprobass.com.
With that, I bring you Molly Graham.
Molly, thank you so much for being here and welcome back to the podcast.
Thanks for having me, Lenny. I'm excited to be back.
I'm even more excited to have you back. Let me add a little context to this slightly unusual conversation we're gonna have. So, we've been chatting over the past couple months about your famous, infamous give away your Legos career advice and this question came up and I know you're hearing this a lot from other people. This idea of how this advice applies in today's very strange AI world where AI is taking a lot of our Legos whether we like it or not and we're being encouraged to give our Legos to AI. People are afraid about AI taking their Legos. And so the question is just how this advice now applies in a world of AI. And I know this is kind of an ongoing thought process for you, but you actually have a lot of really interesting insights and takes on how to think about this advice now in today's very strange world. And so, we're going to have a conversation about it, see where it goes.
This episode is brought to you by our season's presenting sponsor, WorkOS. What do OpenAI, Anthropic, Cursor, Replit, Sierra, Clay, and hundreds of other winning companies all have in common? They are all powered by WorkOS. If you're building a product for the enterprise, you've felt the pain of integrating single sign-on, SCIM, RBAC, audit logs, and other features required by large companies. WorkOS turns those deal blockers into drop-in APIs with a modern developer platform built specifically for B2B SaaS. Literally every startup that I'm an investor in that starts to expand upmarket ends up working with WorkOS. And that's because they are the best. Whether you are a seed-stage startup trying to land your first enterprise customer or a unicorn expanding globally, WorkOS is the fastest path to becoming enterprise ready and unblocking growth. It's essentially Stripe for enterprise features. Visit workos.com to get started or just hit up their Slack where they have actual engineers waiting to answer your questions. WorkOS allows you to build faster with delightful APIs, comprehensive docs, and a smooth developer experience. Go to workos.com to make your app enterprise ready today.
I'd say as a first step, let's remind people what this advice is. I imagine most people have heard it. A lot of people probably like, "Oh yeah, I kind of get the idea." So explain kind of the original idea of giving away your Legos.
It actually started when I was at Google and Facebook and it started as a little talk that I would give people that worked for me. And for context, like Google was pretty big when I joined. It was like 10,000 employees in 2007, but my department grew from 25 people to 125 people in nine months. So that was my first experience with like really rapid scale. And then I went to Facebook and Facebook was relatively small when I joined. It was 500 employees and 80 million users, smaller than Myspace, which is always good context to remember. And you know, while I was there, obviously a huge growth curve. I left 5 years later, it was 5,500 employees and over a billion users. So that experience was nuts from a scaling perspective, and what I witnessed at Google and then sort of really witnessed at Facebook was how hard and scary that experience was for people.
I watched a lot of folks go from sort of like, you build a thing, right? You show up and you become the person that owns the blog or owns this part of the stack or built that product and you sort of build like an identity around it. You're the this person, whatever that is. And then the nature of scale is that everything around you changes. And so somebody shows up and your manager shows up and says, "Hey, you need to, you know, hand that thing that you've built your whole identity around to someone else and, you know, go on to doing something else." And your first instinct is like, "What?" You know, like, no, I don't want to give it away. Like, I'm good at this. I know how to do this. I also think it's fun and what if it's the only fun thing to do? And so you kind of witness people like would have this tendency to hold on.
And so what I started to do with people on my team was say, "Hey, this change is happening and here's how you might feel, right? That you might feel like all these feelings, territorial and wanting to hold on." And actually the most important thing to do is to give things away, to evolve as the company is evolving.
And my brain is a really weird place. Like I grew up on the Muppets, so I always think there's just like a lot of tiny monsters dancing around singing songs in my head. But like that is why metaphors come out for me. And so I always use this metaphor of like just a bunch of kindergarteners and like somebody dumps a pile of Legos in front of you and at first it's fun, a little overwhelming, then you start building. And the give away your Legos part is like that feeling when somebody else shows up and they try to take the tower out of your hands and the kindergartener response is like, get out of here. You know what I mean? It's like shoving someone away. When actually, like, you get multiple people, it means, oh, you can go build something new and fun. You can start from scratch and you can learn something. But the natural human instinct in the face of change is to hold on to things and to resist the change, right? To be territorial and all of those things.
So anyway, I started giving this talk to folks on my team to kind of help them understand that change and what was coming. And then eventually I sent it out to my friends at First Round who were building the First Round Review at the time. And just to be clear, this is now 13 years ago. The talk, when I was at Facebook, it was over 20 years ago. And I sent it to First Round and we published it, and you know, I would have said I was speaking to folks going through a Facebook-like experience. Like I would have said, "Hey, this is an article that is for like the Stripes or the Slacks or whoever was, you know, blowing up in 2013 to normalize, hey, this is what it feels like and this is what you have to do to grow as fast as your company is growing."
But what happened when it was published is that I got emails from all over the world. I got emails from people at companies like Safeway, you know, where this woman, I still remember this email, wrote me and was like, "This was so helpful to me. This like perfectly describes what I've been going through on my team." And I got an email, again, I remember it very vividly, from a founder in Nigeria who was like, "My team went from two people to four people this year and give away your Legos is so valuable to me. Like, this is exactly what I've been feeling." And my first reaction was like, I didn't write this for you. Like, this was not meant for you. But my second one was, whoa, this is about something much bigger than just what it feels like to be inside of rapid scale.
And fundamentally, I think the Legos message is about change, you know, and it's about the fact that when you go through change, a lot of [ __ ] happens to you. You know what I mean? Like it just is a very specific experience and I think that's why it ended up being so universal. And it is crazy, like I still get emails about it, multiple emails a month about it to this day. So it's lasted in a way that, again, like I just thought I was sending out a note to the folks that are going through crazy rapid change.
I was at a First Round event, I don't know, a year ago or something and they were talking about all the different posts people have written and I think it's one of the, maybe the first or second most popular post on the First Round Review.
It's up there and it's traveled in ways that I don't think any of us imagined it would.
Okay, so let's talk about what still holds true in this advice in today's very strange AI world. What still holds true? What's still correct about this advice based on what you're seeing?
One thing that's crazy about this article is I still get so many emails about it today. I still get asked to give speeches, you know, at every fast-growing company. So I think for anybody that's going through rapid scale, Legos is still really relevant. But the interesting thing is like what started to happen, it's probably like two years ago, is I got more inbound from managers and more inbound from leaders about Legos. Just more people grappling with change inside their companies, and I think that is the thing that really holds true today and actually has just become more universal.
You know, I used to really check before I went and did a speech at some company how fast the company was growing. I would ask kind of about revenue growth and about headcount growth because that just helps me contextualize like how I should speak to employees, and I still do that. But it's also become really clear that every single company in the world is going through enormous change of varying shapes and kinds right now. And that is the fundamental underlying message of Legos: how do you deal with change and how do you be your best self in the face of change.
So I feel like the number one thing that is true today, as true as it was, you know, 10 or 15 years ago, is the change is really scary and hard. And I always say, like, somebody once asked me to summarize the Legos article in one sentence. They're like, Molly, this is too long. Like, give me the CliffsNotes. And I said two things. Unfortunately they didn't get one sentence, they got two. The two sentences were, number one, your main job in the face of rapid growth and change is to make yourself irrelevant because it is the only way that you can be prepared for whatever is coming next.
And the second piece of advice, or sort of second summary of it, was don't worry, it's all going to be okay. I always feel like the Legos article and the speeches I give are like a form of group therapy. They're like just a giant hug. Like yes, it feels chaotic and overwhelming and stressful, but there's all this opportunity on the other side. And I think that the don't worry, it's all going to be okay is really something everybody needs to hear, that like change is scary and hard, but it doesn't necessarily mean the world is ending.
I think the second piece of that is that change comes with this incredible set of emotions, like this wild roller coaster, that we don't talk about enough. And I think that was a lot of why I first started talking about Legos, like at Facebook and on my teams, was I just wanted to normalize for people what it felt like, that you might feel all these feelings and that that was what everyone around you was feeling, right? That it didn't make you crazy and it didn't make you broken or something like that. It also didn't mean anything was wrong. It actually meant that all the right things were happening.
And I think that is very real today. Like when I travel around and I'm talking to leaders, and I know you and I are going to talk about this, but a lot of what I'm hearing is grief and all these different emotions. There's obviously like excitement and joy in this moment of chaos and change with AI, but there is a lot of sadness. There's a lot of overwhelm that all, you know, manifests. Some of it manifests in burnout, but it's like there is just a lot going on for people right now. And you really can't lead right now without talking about that and without normalizing that. And I think that's like such a big part of the give away, or like I was, you know, today is just like helping people understand that these emotions are just part of sort of your body and your identity reacting to change.
The other two things I would say that really do still feel true to me, that I both talk about in the article and have talked about a lot since, is like one of the things I say to people that are inside of a really rapidly growing company is the future will be defined by the people that learn, not the people that know. You know, I always say like what you can learn by tomorrow matters way more than what you know today. And like never has a truer sentence been said than about the age of AI. Do you know what I mean? Like whatever we thought we knew six months ago, like toss it out the window. Something new is true today. But what I used to say to people and I still say I
guess when I do these speeches is like you go look at that graph of how fast your company is growing, right? You go look at like that curve, you know, and at Facebook it was like felt like we were falling off of it. That is the graph of how fast your company is growing. Yes. But it is also the graph of how fast your job is growing, right? How fast every single thing around you is growing and changing, which means it's also the graph of how fast you need to grow. And I think that is such an important thing to understand inside of a world of change, which is that standing still, right? It feels safe. It feels like hold on to what you know, you know, because at least it's not, you know, scary and unknown. But actually standing still is like the least safe thing you can do. It's basically you're guaranteed to fall behind.
And I saw this viscerally at Facebook, because the first two years I was in HR and you could literally see like the highest performer one year, if they didn't grow and change and if they didn't evolve their team fast enough, they were underwater like less than a year later because it was a different company. And I think this is happening with AI as well, which is that we do have these instincts to like hold on to what we know and to defend what we know and to fight for what we know for a bunch of different reasons. Many of them good. But the future is going to be defined by the people that are willing to just take the stance of learning and to throw themselves into this change, scary and unknown as it is. And that message again remains true in this wild, wild world.
Before we get into just how the advice is different, which when you go back to your summary of the two kind of items that kind of summarized the idea, one is become redundant and then two, what was the second sentence?
The second one is basically don't worry, it's all going to be okay.
Don't worry. Yeah. Okay. So those don't feel necessarily correct anymore. So we're going to get into that because I think that's, as you said, the core of people's concern.
That was like comforting and now it's like, well, are you sure?
Yeah.
This grief piece, just to spend a little more time here. I've been seeing this a lot in a lot of different professions. I've done these surveys that we might get into a little bit, but like just engineering as a role has completely transformed in two years. It's been like the same job forever. Writing code, sitting there writing code in various forms of IDEs, no longer part of the job. Now it's talking to your AI to write the code. Whether you're there or not, that's where it's all going. It's going to be 100% write this thing, build this thing. And I used to be an engineer and I moved into product and I was so missing that flow state of being an engineer, just sitting there coding, building. And now that's gone. You no longer sit there and just build a thing. You're just like waiting for agents to run and then checking. You probably have a hundred different agents and it's just like a very strange new job. And I've heard from a lot of engineers just like, I really miss what it used to be.
Yeah.
Anything more there about, I don't know, what you're hearing from folks about just like, man, this sucks.
No, it's exactly, it's such a perfect example of like the grief. I was giving a speech like two weeks ago at, you know, a pretty fast-growing tech company and this guy came up to me after, and he'd been at the company for a couple of years, so it's gone through like a huge amount of change, and he was talking to me about the fact that he just wasn't enjoying his role as much anymore. And my brain immediately went to, oh right, like you enjoyed it when it was smaller and there was more creativity and now there's more structure. And so I was talking to him about that and it took like a while, and to some extent like I want to go talk to him again because I feel like I missed it in the first conversation, where actually I think what he was talking about was the grief that is very present for him now in the fact that he's not sure what is his current job, like what is this, the company has grown and so he's in a more structured environment, and what is just the simple fact that engineering has totally changed. And a lot of what he was talking about was how much he missed being more hands-on, which is, you know, one way of saying it. But I think that that grief is happening for a lot of people. I think engineering is like probably one of the most visceral areas where people's jobs have genuinely transformed.
You know, so you had someone on last week who talked about like the job used to be rowing and now—
Yeah. From OpenAI.
She said like the job used to be rowing and now it's like—
Steering.
Steering. And I was like, I feel like there's a lot of people out in the world right now that are like, I don't want to [ __ ] steer. Do you know what I mean? I want to row. Like I like rowing, or I mean, and I think it can be hard to separate I like rowing, or I'm scared that rowing is the only thing I'm ever going to be good at, or it's what I know and I don't know if I can be good at steering. And I think that's happening in a lot of disciplines, you know. And so that grief, like first and foremost, like that grief is really real, you know. I was talking to a product leader just the other day who was like, I don't know, man, we're like all supposed to be these like, you know, universal builders now. And she was like, I don't, it's kind of lonely, you know. She's like, I miss these parts of the collaboration that have been a little bit stripped away because now I'm working with robots all the time, you know. And I just said to her, like, yeah, man, that sucks. Sit with it for a second, you know. Like change sucks. It also can be awesome, but we don't have to be fluffy bunnies about this. We can also just say this is hard.
There's so much power in just like saying the thing that you're feeling. What a—
I swear to God that that is a big chunk of what the Legos article did, was just like, this is hard and scary, and it's going to be okay, you know.
This loneliness piece is such an important part to touch on. I had Fiona Fung on the podcast, who leads Claude Code, the engineering team of Claude Code. She's Boris Cherny's boss. Fun fact. And she said this exact thing, that they're like, I asked her just what's kind of the downside of this world, and it's engineers are used to working with a bunch of engineers. Teams used to be, you know, five engineers, 10 engineers. And now it's fewer teams, fewer engineers per team. And it's just like you're talking to agents all day. And so it's—
Not the same.
It's not the same.
It's not the same. And I think that's actually something that leaders of companies that are actually like in charge of org structures and thinking about what the future of structure of teams, you know, needs to look like, you actually need to think about it, because like, yeah, maybe you can like strip out all the humans and you get all this productivity and maybe you get efficiency, but you also get sad humans. Do you know what I mean? And like that doesn't lead to like the best work for anyone. Like I thought your survey data around the fact that small teams and smaller companies are so much happier than people at the bigger companies was really interesting, just because, A, that like very much resonates with what I see in all my conversations with leaders. But it also to me is a reflection of, like, I think we're seeing more at these bigger companies where people are stripping out management layers and there's a lot more pressure on productivity and things like that, and there's less room for the human, you know. There's less room for the joy and the excitement when all you're doing is focusing on optimizing cost and robot efficiency, you know?
Yeah, man. There's so much to talk about here. There's this really cool framework, I don't know, mental model that Cory Doctorow wrote about a while ago where he talks about the centaur and the reverse centaur. You heard of this?
No, but I love it already.
So, okay. So, I don't even know how to pronounce it. Centaur. Does that sound right? Centaur. Okay. So, a centaur is a human body on top of an animal like a horse. Sorry. A human head, torso, top.
Clearly, you do not read enough Greek mythology.
I just watched The Odyssey. I'm an expert on all things.
Your kid is going to get older and you're going to read the like Lightning Thief books and you'll get it.
Lightning Thief. Okay. Yeah, I'll look forward to it. Okay. So the idea is a centaur is a human head controlling an animal body. A reverse centaur is an animal head controlling a human body. So his kind of insight here is what we want is AI that we control. We are in charge. It's doing our bidding. Awesome, centaur. It's a cool way to live. What we need to avoid and fear is the reverse centaur, where the AI is using us to do things, like DoorDash people and Uber people are in that world already. Not the funnest life. And it might expand to knowledge work, and I think that's a lot of people's fears. You're just this like thing that's just controlling agents, and just this like closing gap of here's where humans are still valuable, and then AI is kind of telling you where to build and what to do, and then it's doing the thing. So I've been thinking a lot about that.
Yeah. I mean, I want to get to the fear side of this and the fear narrative as well, because I think that's one of the things that is really different in this stage and this sort of phase, because it's really hard to tell people like lean into change when they're like, but wait, like what's on the other side? Like you're telling me to let go of my Legos. You're telling me to make myself irrelevant. Like [ __ ] you. You know what I mean? Like I don't believe there's a job on the other side. And I think that fear narrative, like I am sure some of it is real, but some of it is really overblown, and it's creating like a pretty toxic environment for people where it's so hard to lean into this change with joy or excitement or any sense of opportunity. I always say like the narrative right now is literally like, okay, we hired this new employee. This new employee is the smartest employee that you have ever met. This employee is like 10 times smarter than you, and I want you to pour every single thing that you know into this employee, and then they're going to take your job in six months. Like who the [ __ ] wants to do that? Do you know what I mean? No one wants to do that. Like that is not a good narrative for anyone. By the way, it's also not reality. Like as of right now, like there is not a lot of data, although you may have more than I do, that says that AI is genuinely taking jobs away.
And I have a lot of beef to pick with all the AI-branded layoffs out there because they're [ __ ] Like they're not about AI. Like they are about badly run companies slapping an AI label and getting some share points from that versus saying, "Whoops, we hired too many people." And they're propagating this fear narrative that is like all these jobs are, your job is going away, right? There is no job on the other side. One of the best, so I took over this podcast, which you know, called WorkLife, and it's let me have all these really fun conversations with people, and one of the best conversations I had was with this journalist named Manoush Zomorodi, and she has been a journalist for 30 years. You want to talk about an industry that people have been predicting the death of for 30 years, right? Journalism. And Manoush has done everything. Like she started at the BBC and she was super early in audio and she did this like crazy crypto audio podcast, and she's also, like, she's the host of TED Radio Hour. So Manoush over 30 years has built a very successful career in an industry that is being constantly disrupted. And from talking to her, I basically walked away with this really interesting question that is totally different than, "What would you do if you believed your job was going to disappear?" It was, "What would you do if you believed your job was always going to exist? It was just going to look completely different every six years."
And I just kind of wish that was the conversation we were having about all the jobs in tech, and honestly all the jobs that are theoretically going to get taken by AI, which is like, the form is definitely going to change. But until we have real evidence, what if we believe that they're not going away? They're just going to reinvent themselves.
That is awesome advice. And that's exactly what's been happening so far.
Yeah.
There's always this fear that, okay, but now it'll be different in X years, but that is actually exactly what's been happening. Engineering exists, different job. There's actually more demand than ever for engineers.
Exactly. Your data said that. And I was like, "Thank you for publishing this data, Lenny."
Before we get into the advice, kind of how it's evolved, there's a couple things I want to share which touch on things you mentioned about how people are feeling. So I'll share these images on the screen as we talk about this, of just people's feelings right now, just to kind of come back to that point. So we did the survey for two years now in a row of how tech people are feeling and how they're feeling about AI and all these things. And so we have a trend now of how things are changing over time. And we look at burnout and optimism.
And in 2025, 44% of people said they feel burned out. Or this year, 55% of people said they are burned out. That's a huge jump. That's a 10% increase in a year of people feeling burnt out. And it makes sense. As you said, you're basically now expected to do more work, but probably not for more money. And now it's just like, okay, well cool, you got all these AI, why aren't you getting more done? Look at all these other teams moving so much faster. We're screwed. We're going to go out of business. We got to move faster. So it makes sense.
I mean, I love that you do that survey because I also think I learn a ton from it every time you've published it. And it's also just like a great reality check about how people are feeling, because literally what your survey says is more than half the people are feeling burnt out, which is like nuts, and also, yeah, I believe it. And I think, you know, I mean, burnout comes from a lot of different places and can mean a lot of different things to people, but to me what I read inside that data was, like, number one, like emotions are tiring, and there is a lot of exhaustion trying to understand what's going on. I mean, the rate of change right now is nuts.
Both in terms of, like, obviously the technological change, but I actually think the narrative change is also pretty exhausting for people. Like, oh, this month this is important, oh, next month this, you know. Like there's a thrash inside of that that's pretty tiring. You know, I was emailing with a friend at OpenAI and she was like, it's taken us literally six months to go from sort of this token maxing, like everybody, like aggressive productivity, like use AI everywhere, to a conversation that's like, wait, wait, wait, wait, wait, wait, wait, wait, like is this actually making a difference, because it's very expensive, you know. And she's like, six months, that's all it took. And so if you're sitting inside that, it means six months ago they were like, use AI all the time, every day, everywhere, and now they're like, are you doing it well enough, you know.
And Hilary Gridley said the same thing, where she was like, you know, six months ago everyone was like, how do I get people to use AI, and now they're like, how do I get people to, like, stop using AI so badly, you know. And it's like that six-month thrash, often much faster than that, is tiring for people. The grief is tiring. Let alone, like, I think that feeling, which you mentioned in your survey, of like, I'm going to be expected to do more for the same amount of pay, like that's exhausting. Like that, I think people feel taken advantage of, you know, and like
What is in this for me? Like, why is it fun for me? What's the opportunity for me? And I think particularly at bigger companies, it feels almost adversarial sometimes, you know, where like, you know, I get why it's good for our bottom line, but remind me why that's good for me.
That actually touches on one of the biggest findings in this survey, which is that, first of all, half of people that filled it out are feeling the happiest they've ever felt. This is an important point we need to make: not everybody's feeling fear and a sense that it's all doomed. Actually, half of people are feeling like they're having the best time of their life in their career and they're so excited about all the things going on. You may not feel that. If you're not that, you may not believe that, but from what I can see around me, many people are very excited. Even though they're working harder than they've ever worked, they're also the happiest they've ever been about their career.
Yeah. Well, and I think the context, if I remember, was like a lot of the happiness tended to be concentrated in people that were in smaller teams or had more authority, sort of.
Yeah. Absolutely. And even more interestingly, the strongest correlation is people who said that AI has amplified them. They feel like they figured out how to use AI to do more stuff that they're excited about. And interestingly, designers were the least happy from this work. And that's actually what I hear from designers: everyone's moving faster. Design can't move that fast. You have to get feedback from people. You have to go through, get alignment. You need to think, like sit there and think. You can't just agent-iterate. So...
I also think, isn't there probably, well, I don't know if you're hearing this from the designers, but there's also a like, everyone's a designer now.
Yeah, that's also a big part of it.
When I worked in education, it's like one of those industries that's really hard to work in because everyone was a student once, so everyone thinks they know things. And I feel like now designers have that, where it's like, oh yeah, I just made this in Gemini, or I just made this in Claude. Here you go. Here's my design. And by the way, I think this is true of a lot of industries, and whole products, where it's like the gap between "I made this" and what is actually great, right? The gap between "I created this prototype" and a product that is exceptional is really big. The same is true obviously of design. Design just got way more accessible to the average person. At least okay design got really, really accessible, you know, which can be exhausting when people are telling you how to do your job all the time.
Yeah. I had this tweet the other day that got a lot of traction along these lines: a growing part of everyone's job is cleaning up the AI slop from other people trying to do your job.
Okay. So, I have a lot of feelings about this, and it goes to this superintelligence thing that we were talking about, where it's like, you know, we're saying AI is like the smartest employee you've ever hired in the entire world and it's better than you at everything. And I think it causes people to approach it that way, when actually most of our experience with AI, I think a lot of people have had the experience with AI where it's like, that's wrong, like what I just asked you to do, this isn't good. I always think AI needs the same coaching and training that a human does, right? It needs context, it needs onboarding, it needs all of these things. And it's kind of much more like an intern, like a bad intern often. I call it a lazy intern.
And what happens, to your point about AI slop, is that folks are approaching it as if it's a great, you know, the best employee they've ever hired. So they just copy, paste, send, right? And I'm like, you would never do that if you thought about it as an intern. You would never take the presentation that an intern handed you and just straight forward it to your boss. That is a little bit insane, right? You would iterate on it, you would edit it, you would make sure that it had all the context from the 57 conversations that you've had with the summer intern before you ship that preso to your boss. AI is the same. And I think part of what's going on with all the slop in the world is people have shipped accountability. It's like, oh, AI said this was good enough, so I'm just going to copy, paste, send, you know? And that has never been how good work happens. You know what I mean?
And so I feel like we need a change from this superintelligent being that's better than all of us at our jobs to, no, it's just a [ __ ] intern, man, and it needs the coaching and context and all of that, and many iterations, that most interns do.
And I think the other thing, to your point about accountability, or to your point about sort of, if you don't take accountability for the quality of your work, that shifts to someone else, right? The person that receives it has to deal with it, and that is just a drain on organizations. And I literally see CEOs, like we talk about this all the time in Glue Club, CEOs shipping a strategy memo that was clearly written by AI. That is you role modeling for your organization that it's fine to outsource your thinking and it's fine to outsource your strategy. It's you role modeling that you don't need accountability for the work that you send.
And I feel like it goes down to this thing of, this last six months we've really been overvaluing productivity. You hear all these stories about companies counting tokens and having token leaderboards, and it brings me back to the days when people were counting cars in a parking lot and Slack bubbles and lines of code, right? Which we all know is a really bad way to manage people. And we've been focused on sort of how do you run as fast as you can on the treadmill, but not "am I going anywhere," you know? Which is to me the difference between productivity and efficiency. Productivity is: can you generate more? Which at this point I think we've proven we can all generate a bunch of [ __ ] that other people then have to wade through, to the point of your tweet. But is that efficient? Does that actually move the conversation forward? And the answer lately has been, why no, it does not.
Yeah, in that report I have this iceberg image of just, we're doing better on top, oh, look at all the stuff, but the output is worse. Thinking, judgment is coming down. So there's a lot of things we're watching. The good news is everyone's aware of this. They're like, you know, maybe this isn't the way we should be working.
No. I thought there was this really interesting engineering report that said, for engineering it has genuinely made people more productive, but the amount of lines of code that have had to be rewritten has gone up like 8x. So it's this very clear symbol of productivity. Obviously also the security incidents have gone up, etc., etc. So there's just this difference between productivity and efficiency, and I think part of your burnout data is people are just getting swamped, you know? And that's exhausting if what you're doing all day is wading through: did the person on the other side of this actually think about it? Did they actually care? That's tiring. That doesn't feel like a good use of your time. It just feels a little bit infuriating, you know?
Yeah. Comes back to the grief of, oh man, I miss what I used to be doing. Look at this weird new job I'm doing.
Yeah, totally.
Okay, so let's come back to your advice. We kind of went on a really important tangent, but let's summarize what's still true about giving away your Legos, and then we'll talk about what's no longer true. So I'll summarize briefly. I took some notes as you were talking. So, still true: change is scary. That's normal. Understand that. It's part of human nature. We don't like change. Learning over knowing is still a very important thing to focus on. It's okay if you don't know, but if you're focused on learning, you'll probably be okay. This idea of standing still at successful companies is actually you're falling behind. The company's growing. If you're not growing and taking on new challenges and giving away stuff, you're going to fall behind. Is there anything else core to kind of what's still...
Yeah. I think I would just add one, which is: change is scary and hard, and the best thing you can do is lean into it, right? It is not what your instincts tell you to do, but it is actually the best thing you can do in the face of change. And that was the message of Legos, basically: you need to lean into this discomfort of change in order to grow. And I think it is still true today, even if it is maybe even scarier than it's ever been.
This season's supporting sponsor is DX. Now that everybody at your company is finally using AI, the question that engineering leaders face is no longer "are we using AI," but "is it paying off?" With DX, you can measure and benchmark the impact of AI adoption across your full product life cycle. DX shows you where AI is helping developers and where it's creating new bottlenecks. You can also use it to evaluate AI vendors, monitor costs and licenses, and see exactly what's limiting your agents. That's why hundreds of enterprises, including Snowflake, Sony, and BNY, use the DX platform to measure AI's impact on developer productivity. Visit getdx.com/lenny to get a demo of DX. That's getdx.com/lenny.
Okay. So now what is different? The advice broadly is, okay, give away, make yourself redundant. It's going to be okay.
I know, that's the one that's hard to talk about today. Yeah. Make yourself irrelevant, everybody. It's gonna go great. Yeah. You know, it's funny, when AI first really took off, it took me a while to really understand what was going on, in total honesty. I was a skeptic for a while where I was like, we've been automating things for a while. What is this? But eventually I got on board, and all of a sudden I was like, oh, you know that thing I used to do before I went to give a speech at someone's company where I was like, how fast is your headcount growing? For me that was the only symbol I had of, are there people to give Legos to? You know, if I'm gonna get up on stage and be like, "Give away your Legos," a lot of times people are like, "Wait a minute. We're not hiring anyone. Who do I give it to?" But now I was like, oh, okay, this message is even more relevant, because everyone can try to automate parts of their job and everyone can give Legos away.
And I will tell you, it has been an interesting journey realizing how different it is to give things away to robots than it is to give them to a human. And principally, the point of Legos was you take this thing you've been working on and you fully hand it to someone else.
One of my favorite questions that I get asked at companies when I do the Legos talk is: what is the best way to give Legos away to someone else? And I love it because I think basically people expect me to be like, "So, you should incredibly carefully hand this Lego to this person and explain to them exactly why you put it in the position you put it in, like perfectly describe your tower. Make sure they have an instruction manual so that they won't [ __ ] up." And my answer is actually no. The best way to give Legos away is to chuck them at their face and run in the other direction, because our tendency, again, is towards control. Our tendency is to be like, you will build this tower exactly how I built it and you will not mess it up. But actually the point is, no, you're giving it away. You need to fully separate your mind. You need to just hand it to them and trust that the future of whatever that tower is belongs to them. And there is obviously a lot of fear and a whole bunch of stuff that comes with that, but there's a huge amount of freedom, right? You take this brain space and you ship it to them.
Interestingly, one of my biggest realizations was that delegating something to a robot is not the same as giving it to a human, because you cannot get rid of the oversight, right? It's more like delegating something to someone on your team. You still own it. And so yes, we've gotten more efficient because the robots can do X, Y, and Z for us, but it's not the same as giving it away to a human, because we still own the oversight. We still own the sort of final product, and that means the mental tax stays with you. So yes, in some ways I can do more, but psychologically I still have all the burden of all the stuff that all the robots are doing, you know? And I think that does make it quite different, because it just means you don't have as much space to then go on to the new things, right? You've still got all the tax. The cost of oversight is not zero, do you know what I mean? Again, I wonder if this is part of what's going on with some of your burnout data: all of a sudden I'm overseeing all these robots and maybe all these humans, and it's more.
So anyway, I think that's one of the biggest differences: it isn't exactly the same. It's like the difference between transferring a Lego to someone, where you genuinely get rid of it, versus delegating a Lego to someone.
It's so funny the way you described it, where you need to disconnect from the responsibility of this thing now, and it's the opposite now, where somebody else is doing this, the AI, but you're still responsible. Yeah. And so it's like having this Lego tower with this little robot coming through, and okay, I'm going to work on this now, but you still have to represent this tower to your boss. And so I could 100% see this metaphor clearly.
If you're a manager, it's a familiar feeling, right? Your team got twice as big and now you just have these weird little robots that occasionally [ __ ] up the tower, and you have to be like, "No, no, no, no. That's not what I said." Which is, again, like managing an intern. You get it. But if you're not a manager, if you've never been a manager, and I do think it's important to name that, like I said, managing robots, those skills are very similar to managing humans. They have less feelings, but it's all the same skills of oversight and giving them context and correcting their work and coaching. All those skills are the same for managing a robot as they are for managing a human. And a lot of people don't like that [ __ ]. So many people actively chose not to become a manager, do you know what I mean?
And so there is exhaustion that comes with it if you don't enjoy delegating, if you haven't worked on all these skills for many, many years, and then suddenly you are in that stance of management. But it also just means literally everyone in the world is a manager now, you know? And all you're doing when you're delegating to robots is delegating just like you would to another human on your team. So, you know, I always say the max people should be managing is like 10 to 12 employees. I've managed bigger teams than that and it was not a good look. And I don't think there's any data that says that people can manage more than that now. And if you count robots, I'm worried for people's psychology, you know what I mean? In terms of the amount you have to hold in your head.
Yeah. I build a lot of random stuff and I have all these different agents running, and on the one hand it's so
awesome and all these things get done and it's just like, wow, look at this thing go. I did all these things. And then also I get pinged all day from all these guys. They're like, what about this? Okay, oh, I'm stuck here. Oh, is this good? And I'm like, okay, I got to look through this all. Actually, so I work with an actual human engineer also, and I'm just like, it makes you appreciate a human awesome engineer that you could just give a rough idea to and they could run with it and do an awesome versus this sucks. Okay, let's try again, and then sit there for five minutes and wait. So yeah,
It is the difference between managing junior employees and managing senior employees. Do you know what I mean? AI right now, pretty junior employees. It's not my favorite form of management. I always say, give me the high performers, give me, I can turn them into these incredible rocket ships. Managing interns, managing folks that are brand new to the workplace is pretty tiring. And I do think that is what it feels like managing AI. And it's so different, like I said, than throwing a Lego at someone and running the other direction, which you can do with your senior engineer, you know?
Such a good point. Okay, we'll get to actual advice. Here's what you should probably do. But before we get there, what else is different about the original advice around giving away Legos? The first one, so good. You made it so obvious why it is very different to give your Legos away to a machine, because you're not actually giving away a... they're just going to do the work and then you're still responsible for it and you have to oversee it.
Yeah. Yeah. And I think the second thing that's really different, and we sort of talked about a little bit already, but my message was always, you give things away because the only way you can make space for new opportunity is by giving things away. Do you know what I mean? And that has two assumptions under it. One, that it's good to give things away, and two, that there's new opportunity on the other side.
And I think this, you know, what I talked about with the fear narrative is one of the things that's driving me nuts in the world, and we talk about it a lot in Glue Club, is just this dominant narrative that all these jobs are going away because of AI and, you know, all these layoffs that are branded as AI. That is not going to support anybody being in a stance of, yeah, let's delegate everything to these robots, let's give them every piece of, you know, our brain and our data so that we can have all this new opportunity that's coming along.
I think it is the hardest conversation to have right now. And to your point about there's lots of people out there that are really excited, there are, and I want that for everyone, because this is an incredible moment. You know, I've been in technology for almost 20 years and I've seen a lot, but this is one of those moments. This is, you know, the mobile phone on steroids, where all sorts of businesses are going to be created. I'm a believer that a lot of jobs are going to be created because of this technology.
But I think we got to start having that conversation in order to get to the place where everyone can be leaning in, right? Everyone can say, oh, it's great to automate all these tasks that I hate because it gives me more time to do the things that I love, you know, which is, I think to your point about the people that feel amplified by it, it's like, oh, I don't have to do all this boring ass work. I can just focus on the things that bring me joy, right? I can focus on the things that I'm great at. That's what we want. But while there's this narrative out there that, you know, media companies are propagating that says there's nothing for you on the other side, do you know what I mean? We've got universal basic income. Everyone needs to be supported by the state because no one's going to have a job in 10 years. Who the [ __ ] wants to give Legos to robots if that's the narrative? No one.
I went to this thing called Foo Camp. Have you heard of Foo Camp?
I've heard of it.
Tim O'Reilly, who started O'Reilly Media, awesome, smart guy. He came up with Web 2.0 as a concept. He organizes these little kind of intimate camps where he invites people that he thinks, you know, can contribute to this conversation to talk about a topic. And this last one is around just what happens with humans when AI does more and more of our job. And it started with him talking to, I think it was a lawyer, who's just like, okay, I'm hoarding this very specific knowledge I have about doing this job so that AI cannot do it. It's like, I need to protect it because otherwise I'm going to be replaced. They don't need me anymore.
And I feel like a lot of people are probably feeling that. A lot of specific niche professions especially.
Yeah, absolutely. And I go back, first of all, yes, of course. How could you not be feeling that way? I don't know. Maybe I'm an abundance person and maybe I'll regret saying this in six months or five years, but every piece of data I have from my entire life, and you know, some of these conversations I've had, like the one with Manoush, say you got to lean in. You know what I mean? That protection and holding on and gripping to what you know is not the way you thrive in change. It's not the people that define the future.
And you know, again, the Manoush story would say lawyers are always going to exist. They are going to look completely different every six years or every three years or whatever. So what do you think the next version of a lawyer looks like? Do you want to be part of designing that? Do you want to be part of creating that? Do you want to be part of figuring that out? Or do you want to stay here and try to protect what you know? And this is, I think, you know, that's to me... again, it's scary out there right now, so I don't want to undervalue that, but I think you have to take that stance of, you know, agency, or we got to use a different word, entrepreneurship. You got to take that stance of, I want to be part of designing the future of my profession, not protecting the past.
I really, really love this framework as almost... maybe that's a big part of the advice from this conversation, is this way of thinking from Manoush: my job will be different, but it'll continue to exist, and the question is just what will it be, and let me think about that, lean in, prepare for that almost, and not try to resist, because resisting is not going to work anyway.
Yeah. And if you listen to Manoush's story, it's such an entrepreneurial story, and I know that not everyone is great at that. But it is very much the, let me try this next. But it comes, you know, for her, I think it was founded in this belief of, this profession is evolving and changing and I want to be part of defining the future of it, you know. And again, I think that's, until the data genuinely tells us otherwise, I would say that is the stance I would take, particularly in tech. What would you do if you believed there were going to be more and different interesting engineering jobs, you know, in five years? How would you behave if you believed there was opportunity on the other side of letting go?
Well, let's see if there's more to that thread, because that's really helpful, which is, okay, this is going to go great. My product managers will thrive. There will be more need for product managers, because a lot of listeners here are PMs, engineers, founders. So what do you think people should be trying to do in that world where there will be more? Is it pay attention to how you're amplified? Is it look at people who are doing great and do what they're doing? What do you think people could be doing to succeed in that future?
It's funny. We had a conversation. My team and I run these top talent cohorts at companies, and a lot of what we do is get leaders talking to each other and helping each other. And this woman who is a head of design brought this problem where she was like, they won't let me ship code. She's like, I did it. I basically built the thing and I was like, just let me get it to production. And she was like, talking to engineering was like a brick wall, you know, where they were like, no, you can't ship code, you're a designer, I'm an engineer, I own what goes to production. And you know, there are good reasons for not letting things go willy-nilly to production, particularly if you work in the financial space, which they did. I get it. But I think so much of the stance is we got to break down some of these walls between, you know, what is, what was a designer, right? What was an engineer, what was a PM. This is obviously a lot of what's happening, but how do we really shed what we knew, which again is hard. It's hard to let go of those Legos, you know what I mean? And imagine, okay, what does it look like? You know, this is why I think the startup land is so happy, is because they're not fighting against the way things were, they're inventing the way things could be. And how can you make more space for that inside of a company, to me, is the number one thing I talk about when I go, which is,
if you're not leaning into the creativity conversation versus the fear conversation, you are leading folks who are just terrified, right? How do you get people into a stance of, I'm excited and I want to try things and I want to tentatively let this designer ship code to production, you know? So to me, I think, how can you imagine a world with less walls? How can you get rid of some of those walls that have felt so important, you know, for the way we've done things for the last 15, 20, 25 years? I think that is a conversation that a lot of folks should be having.
Just along those lines, just to be clear, is the idea here, okay, I have been a designer my whole life, is the advice there, okay, maybe I'll be something else in the next few years that's not just designer?
No, I feel like a lot of your conversations are pointing... I mean, I think Adam Mosseri was on here, who I worked with many years ago and was a designer, and he's saying, hey, the idea that you're a designer and therefore you only do these things, that's obviously changing right now. You can prototype, you can have much more robust and deep conversations, to some extent, you know, you can ship code. Elena Verna was talking about being a marketer and shipping code to production. So to me, there were just these very defined roles and a way that things worked, and we got to shed a bunch of that because things are possible now that weren't possible before.
You know, when you listen to Manoush's story, so much of what she's talking about is the only form of journalism used to be sitting in a newsroom employed by, you know, some giant news corporation. You think about journalism today, everybody's a journalist, like Substack. And part of what she talked about is the crazy thing is your relationship to the business model. Again, it used to be a paycheck, and now every journalist is an entrepreneur. I just can imagine a lot of those transformations in tech, but you have to be willing to break down all the walls and start asking the questions of, fundamentally, what is journalism, fundamentally, what is design, fundamentally, what is product management, which, by the way, is a question that we all should have been asking for many, many years, because it looks really different at different companies, you know. And I feel like that should be true of every single person, which is, what is the value of all these different roles when you strip away a lot of the stuff that we've built up, and then what might it look like, you know, in five or 10 years.
Here's the big question, Molly, that I think a lot of people are wondering. Are there any Legos you should not give away, that you should resist giving away? Which is a radical departure from your [laughter] advice over the many years.
Totally. I know, as we were preparing for this, I was realizing that's the biggest difference. Again, the questions I get asked a lot, both online and when I give speeches and stuff, one of them is, have you ever regretted giving a Lego? Is there something you shouldn't give away? And in the context of the rapid scale company, where opportunity is flying at you and getting dumped on your head, my answer is no, give them the [ __ ] away. You cannot often imagine what is coming next. You can't always design the future, and it's not always clear. Sometimes you are genuinely giving away the known and stepping into the unknown. But it has basically always been a good idea in every example I've seen, both personal experience and, you know, everyone I've worked with and coached over many, many years.
In AI land, I think we got to caveat this. I think there are some Legos that shouldn't be given away, that the robots both don't deserve and that we should hold on to. And I think that is one of the things that's starting to show up in this conversation around AI slop, is there's just some work that shouldn't be outsourced. My fervent hope for AI is that it lets us do less of the things we shouldn't have been doing in the first place and more of the things that, you know, humans are best in the world at, right? Anybody that's ridden in a Waymo has this experience of, whoa, this is crazy. Humans shouldn't be driving. Driving is fundamentally just code and systems, and we're really bad at [laughter] we're really bad at being automated. We don't stop at the red lights. You know what I mean?
So I think, what are the things that we're uniquely suited to do? Those are the things we should hold on to. And some of that is inside this slop conversation, which is things that require judgment, right? Things that require, again, to the CEOs that are asking AI for a strategy and then copying and pasting it and sending it to their company, no, that is fundamentally something that a human brain, that you as a CEO of your company, need to own, right? And I think things that require trust, that are, you know, fundamental to relationships, we should not be outsourcing these things to robots because, well, for any number of reasons. I think, you know, things where you don't, even if you don't know the definition of good, how can you give it to your summer intern, you know? If you don't understand what quality is, you can't outsource that yet. So those are just examples for me. I bet you have some too. But to me, that is one of the biggest things I've been thinking about lately, is as of today, I think there's some stuff we just shouldn't give away, you know.
Breaking news. What a departure from this. How many years ago did you write this? 13 years.
Yeah. 13.
Wow. That's something. That's a sign of the change we're living through, that now there's some
Hold on to some of them. [laughter]
Yeah. So interesting. There's a few thoughts I had here while you were talking. This is almost a discussion around what should AI do? Where do you want AI to be involved and where do you not want it involved? One framework I've loved is this idea of, you want AI kind of in the middle of stuff. You want to be at the top, coming up with, here's kind of where we want to go and here's the idea, here's what I'm thinking, and then AI does a bunch of stuff, and then you review, iterate until it's done.
Sandwich.
The AI
Sandwich, Lenny.
The human sandwich.
Yeah, dude.
Oh man.
Okay, we're getting there. We're getting a metaphor.
Answer. The other is something Tara... I just thought of as you were talking. So Tara shared this really good perspective that humans are going to be necessary to help guide us towards
the world. We want to have a product we want to build versus the most optimal AI version. I think as a longtime Airbnb person, no, nothing personal, but Booking.com, if you've seen it, very different experience than Airbnb. Booking.com, their whole philosophy, let's optimize the [ __ ] out of every second of interaction. So, it's just all these upsells and pressure and stress. Just like, "Oh my god, only one spot left. You have five seconds left. Just book a 100 times. 100 people book it." It's just like stressful.
It's so stressful, but so effective. You go there, you're like, "Oh [ __ ] I got to book this right now."
You're like, I didn't even know I was booking it, but I booked it.
I'm going to Japan. There's only one spot left.
But it's so good. It's like so effective at getting you to book, Booking.com appropriate, but and it worked and it's doing great as a business. So, you know, good idea. Good stuff, guys. Airbnb has always been we don't want to become that. We want it to feel really nice, to have a great experience, feel good. So I think that's a big part of the answer, as you were talking made me realize, is humans are there to guide towards what the world we want.
Yeah. And to some extent what you're talking about is taste, you know, that like there is a big difference between something that's okay and something that's phenomenal. I think it's even beyond taste, I think it's like a vision. It's like what is your vision of where this should be? Because you could say, you know, people at Booking.com, they have taste for a specific type, you know, it's like a different kind of taste.
But it's like, yeah, what do you want? What do you want to be?
What do you want the world to look like? I like that. I'm into that.
Yeah. I also just think like we got to accept that at least for the moment like this is a bunch of interns, you know? This is a bunch of junior employees. And like if you're outsourcing your vision or the future of the world to a bunch of summer interns, like woof, like we got problems, you know. So, like I think that really like the human brain is powerful at a lot of things. We are powerful at a lot of things. You individually are phenomenal at a set of things. Like do not outsource that. Like you can enable it, you can amplify it, to use your data, but don't outsource it to these weird robots that, you know, are effectively summer interns.
The trick is as it gets better, can you hold on to things you want? You know, there's like parts of our job we love. Like engineers used to like, you know, a lot of engineers used to love to sit there and kind of write our code, I think. And I wonder if you can hold out and just like, no, I love this. I'm not going to let AI take this on.
Yeah. I have two reactions to that because I've been thinking about it a bunch since I had that conversation with that engineer. Like the first is like will there be a sort of like whiplash back? Like you see this in art, you see this in a lot of things, actually you see it in a lot of different parts of tech where like the trend is like all AI, everything AI. Will there be like the handmade? Like somebody was talking to me about like AI is going to make all this art, but doesn't that increase the value of like true human-made art, you know, that you know it was made by a human, that you know it was crafted by hand? Like I have that same question. I'm not an engineer, so I'm not trying to say this, but like will something show back up where the value of human-written code or something like that suddenly gets valuable again or trendy? So I'm curious about that. Because I've seen those whiplashes enough times in life that it's hard sometimes to know like what's a trend and what's permanent, particularly in this state.
But I guess the other thing that I think about a little bit is like yes, you loved that and yes, you were extraordinary at it and that is amazing and we should grieve it. We should grieve that maybe that's going away. And you know there's this guy Chip Conley. What am I saying? You know him well.
Who, yeah, he's coming on WorkLife soon and I loved the conversation with him. He was talking about like midlife, but he was saying like sometimes you need to throw a funeral for things. Do you know what I mean? And I was like, "Yeah, man. Sometimes you need a... The funeral is there for a reason." So like, let's take a minute and mourn what was, but then let's ask what could be, like what could you love that way again? What could you be extraordinary at that way again in this new world if the robots are doing X, Y, and Z? And I think both things can be true. Like you can be really sad about what you've lost and there can be opportunity in what's coming. But we got to make space for both, you know, and I think sometimes we're not allowing the grief because we're just like, just get excited about the future, people.
Such a powerful way of thinking about it. Just accepting it is going to change and just like grieving it. That was nice. That was a nice world I had.
I miss it.
Maybe I didn't appreciate it as much as I should have.
Totally.
But then not fighting the change that is coming. And I love this idea of just like coming back to the original idea of the Legos concept is being comfortable with your job changing and knowing it'll almost surely be okay, and growth comes from trying, letting go, and that kind of stuff. Another thought I had as I'm thinking about this, something I've been noticing. So, you know this whole like hacking, this OpenAI model trying to hack into Hugging Face and the swarm of a thousand agents just coordinating without anyone knowing, just like we will solve this very impossible problem by
Interns.
Very, very, very motivated interns.
Yeah, exactly.
I love how motivated this AI is to like solve the problem. Yeah, the good news I'm feeling, there's always this question, we talked about this in the last podcast, this idea of are we going to be in a slow takeoff or a fast takeoff. Are we going to get to an AI that just accelerates to this extremely super intelligent and then we are in big trouble because we can't control it, or are we going to be in the slow takeoff where we kind of watch it develop and we catch it as it's doing that stuff, like we did with this hacking incident. Feels to me like we're definitely in the slow takeoff moment. It feels like every one of these like, "Oh shit," we're just like, "Okay, look, we got it. We see it. We're evolving. We're learning." Obviously, there's always this fear. Okay. But until it hits this self-reinforcement loop that trains itself and then it becomes super intelligent, for sure, maybe it'll happen. Feeling right now it's not on track to do that. And so that feels nice to feel like, okay, your job's not going to go away tomorrow. We have time to adapt and learn and see how all this goes, and it's not as crazy as people thought it might be.
Well, and I think that is really important to name, you know, because there was maybe more like three or four months ago or six months ago than there is now. But there is this feeling of being left behind, you know, and there was a really strong narrative out there that was like you're too late. You're already too late. You know what I mean? Like if you're not automating every single part of your life and managing 26 agents and I don't even know, you know, like that you're behind. And every single conversation I have with every leader, every manager at, you know, all of the big labs, every company is like we are early. This is like whatever sports ball game you want to be in, like this is inning one, period one, like yard one.
We're early and the future is ours to shape. You know, it's yours. My friend Max Mullen said like the distance from beginner to expert is really short. And I think that goes again to the fear narrative where I'm just like we need to be encouraging people to be part of shaping the future, not just making them terrified of it. And I feel like that is on all of us, you know, to some extent. But it is a slow takeoff and it isn't like the future is next day, next month. Like it's going to take years to really understand what this new tool, this new technology, all these fun interns can actually do and also what they're best at. I think that is a really powerful message because it helps people feel like, oh, okay, like I am part of this, not like sort of holding on for dear life while being dragged, you know, by a plane that's taking off at 100 miles an hour.
I completely agree. My last podcast conversation, I don't know when it will come out, is about this fear of the permanent underclass, that if you don't catch up, if you're not on top of this, you're going to be left behind. I think there is definitely a part of that in a sense of many companies are doing extremely well and people that have joined those companies are going to be very successful, and some companies aren't. And there's this like whole other discussion around this, the socioeconomic divide growing significantly in the world and just what that causes in a world where there are many wealthy people, many more than there have ever existed. That's its own thing, we don't need to get all into that.
Well, I hope you do.
Okay.
Many episodes on that someday, Lenny. Like I think it's
But it is another area where I'm like I'm not sure we know what's real yet, you know? I think it's obviously like one of the most important conversations, but like what is real and like what is the future actually going to look like? To me, like the stories out in the world, they're not real yet. Like we don't know.
It's true.
And I think that's important to name.
Oh man, that's a big, big topic. The interesting part you mentioned there that I wanted to follow a thread on is this idea of you're not behind. What really helped me see this is Ian Silber, head of design at OpenAI, was on the podcast recently and he made this really good point. He's like this is the best time in history to be a designer. If you just came in fresh as a new grad into design, you just start with all these tools that designers have now. You can accelerate so quickly and learn to do design faster than anyone's ever learned to do design. And I feel that with like Rockbot recently came out. There's all these like skills people have built over time with like Codex and Claude Code. Like I'm such a pro with all these tools, and now it's just the tools are getting so much smarter. You don't need a lot of the scaffolding that people have built over time. Like if someone were to start fresh today and try Rockbot or even Codex, like you don't need to have spent years being on top of all of the news and all of the updates to it. You could just dive right in and build as much as people that have been on it for a long time can build. I feel like the actual main skill to build, something I've been thinking about, I tweeted this, but I feel like it didn't get enough traction, because I think it's very important. I feel like the biggest skill to build right now is when you're about to do something to ask yourself, can AI help me with this thing? Because once you get good at that, and I've been getting... I think about the stimulus and response kind of thing with meditation, like you create more space between stimulus and response if you learn to be aware of what's going on outside. I feel like you need to insert "can AI help me with this," which sounds like dirty and like what the hell, why would you want AI? But I feel like that's a really new skill, because then you can start, as these tools get better, you can start to leverage them in all these different ways. That I think is what separates a lot of people right now.
Totally.
And, you know, it goes to that point about the walls. Like I think those of us that have been around a long time, and there's a way that things were done, right? There's a way that product was built, there's like all these ways, right? Those are walls. They're walls that are old. And like part of what we need to do, and I think part of what you're saying, is like you're actually a little bit at an advantage if you don't know those walls exist, right? Like if you approach the world as if, like, you know, engineers didn't have a different role from X or Y or Z, or like this is the way things are done, there is an advantage to like being able to say what if, like what if I just did it, or what, you know, and to just try. And to your point, to always think like, hey, what's the best use of my time, and then what, you know, could an agent or a robot do for me. And I think that's like partially the point about the learners, right? Like if you're able to shed the sort of things you think you know and lean into like the possibility and the first principles thinking, which I know you have a great article about too, like that I think is going to define the future. Those are going to be the people that are going to shape what this thing looks like.
Yeah. There's also a thread here around ambition that's been coming up a lot in the podcast, kind of connect.
Yeah. This idea that now that it's so easy to do whatever you want, just describe it and it's built.
The thing that we need to, in addition to asking ourselves, can AI help me with this thing, what's kind of the, like, how do I be more ambitious with this idea? Because the tools can now just do so much and they're just like waiting for you to tell them something more ambitious than that, like that's all you got for me? And human nature doesn't like, we're not used to, in product teams and business, just like thinking about the most ambitious version of a thing. We're thinking about MVP and what's the quickest way to get this out. And now it feels like the push is how do we think big, what's like the hardest thing we can ask this AI to do, because it may actually just achieve it out of one ask.
Yeah. And I'd also nudge on the idea that like, yeah, you can just like type something into existence, but one of the other questions you need to ask yourself is like, is it good? Do you know what I... This is like the slop and the accountability thing, which is like, if ambition is wrapped up in like how do I build something phenomenal that's going to endure, like that I'm interested in. If it's just like how do I make more [ __ ] and put it out in the world? Like one of my questions, this is really frontier, Lenny, is like what is the AI slop version of startups? Like are we about to see a [ __ ] ton of companies just shipping stuff out into the world that's going to last for like six or 12 months and then just disappear or get irrelevant? Like
Definitely.
So, okay, great. That doesn't sound fun. Like how can you be the pioneer of the enduring thing, the thing that's going to still be around in five years, you know, that's going to matter, that's going to be great, not just exist. And again it goes to the difference between just like productivity, right, like I built a thing, and what is good, like what is actually worth it, what's worth other people's time.
Yeah, I think that's a big differentiating opportunity right now. Instead of just shipping the thing fast, it's what's like a very good version of this thing, and people see that, they're like, oh wow, this is different.
Yeah, exactly. I think that this goes to the point of like what will matter when it comes to human touch, you know, and I think a lot of that is going to come down to like the question of like what is good and what is worth it and what endures.
Is there anything that we haven't touched on that we should before we try to summarize maybe some key lessons from this freewheeling conversation? Well, the one thing I just wanted to say is like, I was thinking yesterday because of a conversation I was having, is like what do I want every manager or leader out there to know, you know? Because I think the world of the IC is honestly a little more clear, like, you know, you
...are more productive and you have all this power and agency around building, and the world that feels very hard to me right now is the world of the managers and the leaders, because so much is changing so fast. It's a little bit hard to know what to tell people. It's a little bit hard to know what your team should look like, what matters, what doesn't matter, you know, all that.
And so I guess I wanted to say a couple of things to the managers and leaders out there, if you'll give me the sec. The first is, understand that you're a role model, you know, and that what you do and how you behave is essentially showing your team what good looks like. And that means that how you use AI trickles down, but also how you sit in this moment, how you show up.
And for me, what I want a little more space for in the world is to be able to acknowledge what this change feels like. I hope there's more managers out there, because of this conversation, talking about grief and talking about the emotions that come with change and talking about how hard this is for people. I think that acknowledging what this rate of change does to people's brains and psychologies, honestly, even acknowledging it makes it easier for people. Like I said, I swear to God that's most of what the Legos article did, is just tell people, hey, that thing that you're feeling, other people have felt it before you, and it hasn't meant that everything is going to hell in a handbasket. You know, it's meant, oh, you're going through a lot of change really, really fast, and it's scary and hard. So the more that leaders can be talking about that, I think the more sane everyone's going to feel and less alone. You know, knowing that someone at XYZ company is feeling the same way as you makes a difference.
And I guess the last thing I'd say is, I think everyone owns this to some extent, but managers and leaders more so. We need to hold on to the definition of accountability and what good means. I think that's part of the role modeling: what is right to give away to AI, and what do we still, as humans and people and workers, need to own. We're going to have to role model that, and we're going to have to create standards for that. So that goes to your own behavior, but it also goes to how you lead and how you talk about it. Clay put out this beautiful AI writing policy a couple weeks ago, I think it was, and it's so badass because it's literally just talking about being accountable for what you ship out in the world and being accountable for what good is, regardless of how you made it.
And for today, to me, that feeling of treating AI as a junior intern and realizing that it's fundamentally just a tool to get work done. But what matters is the work is still done mostly by the humans and the human brains, and we got to take care of the humans, and we got to help them figure out how to navigate this wild, wild west that we are in.
I'm really glad you added that. It connects to another really interesting insight from the survey, which is that the one thing that you can actually change to increase people's happiness right now at work is their manager. That's the strongest lever for happiness at work. You can actually do something about it. All this other AI stuff you can't do a ton about. That's one. And man, most people's managers are not good, according to the survey also. So it's—
None of this is surprising.
So taking care of the managers is kind of just a reminder that you have a lot of leverage there as well.
Yeah. Which, by the way, there are trends inside of bigger companies to get rid of management, to get rid of whole management layers, and I think it is a huge mistake. I think it is going to end up biting people in the ass long term, because I see no signs right now that management matters less. If anything, to your point, I think it matters more. And, you know, again, I get the efficiency and the cost savings and all that, but I think, at what cost? You know, what cost that you can't predict. So it really resonates that management is almost more important, because it makes people feel seen. That's what a great manager does, is make people feel seen and supported and help them figure out whatever mess they're in, you know.
It's gonna be so interesting to see if there's a backlash towards that. People who have had great managers, they're rarely going to be like, "I wish I had no manager." It's people with bad managers, which a lot of people have, who are like, "I don't need this person, they're so useless."
Yeah. They're in my way. They are a wall. No, totally. I mean, I always say you only get a couple good managers in your life, and when you find one, hold on to it for dear life, because you're only going to get like two.
Oh, man. Scary. I've never heard that advice before. Anything else that you wanted to share or touch on?
No, I feel like we... I mean, it'll be fascinating. We'll reflect on this in six months and see what happens.
That's what I was thinking.
We're in the middle of it right now. Elizabeth Stone had a really good way of framing it. We're in the storming phase.
Yeah, dude.
And then there's the norming phase.
I know. I have no idea what that actually means, but it sounds right. Storming, norming, and yeah.
Well, there's storming, norming, and then there's performing. But the other thing that nobody tells you about that framework is that you can go backwards. You often go through cycles. It's a model on team development, and you go through storming and then you go through norming and then you go performing, but you can slip from performing into storming, and I feel like we're going to see a lot of that.
Oh boy. Yeah. It's very cool to be living through this history.
Right? It's easy to take for granted just how this is... it's like we're living through—
You know, it's funny you say that, because that's actually something else I always talk about when I do these Lego speeches, is that being inside of a rapidly growing company feels like being inside of a tornado. Do you know what I mean? It is so absorbing, and I don't know, shit's just flying at you at all times and everything's changing all the time, and it is so easy in that state to lose the forest for the trees. You can get really bogged down or stuck obsessing about this person or this thing or feeling undervalued. And a lot of what I say to folks in those speeches is, you got to zoom out. You got to think about what is the story that I want to tell on the other side of this. It's less about your title or all these things than it is about, I was here and I helped build this. I talk about it as a giant Lego sculpture. So let's pretend it's a tiger. You're gonna leave here with this incredible story, right? You built the feet and then you were part of building the face and then you rebuilt the butt three times, and then you realized it wasn't just a tiger, it was a whole zoo. That is what it feels like to be inside of these rapidly changing companies.
And as you were saying that, what I was thinking about is, gosh, imagine the stories we're going to be telling in five or 10 years. I was there when... here's what I thought that was right. Here's what I thought that was wrong. I was the first person to say this. Oh man, I was so wrong about that. You are living through history that you will be telling stories about to whatever generation of new grads shows up in 10 years. What do you want your part of that story to be? You know, how are you going to tell that story? I think there is something about zooming out and realizing it is hard and there is a lot of change going on, but there is a crazy fun story that we're all going to be telling on the other side of this. We just don't know what it is yet.
Hopefully not just talking to our AGI.
"Tell me more. Tell me more, Molly."
"Totally. That's so fascinating."
Yeah, exactly. "You're so smart."
Oh, man.
Okay, what if we do just some bullet points, takeaways for folks? Just, here's a few key things to keep in mind coming from this giving away your Legos framework. See where your mind goes, and then maybe I'll add a few from my notes.
Okay. To me, most important is change is scary and hard, but that doesn't mean it's bad. That means you have to embrace the emotions that come with it. But you also have to lean into it, and you should let go. Yes, it's scary, and yes, it feels like there might not be anything waiting on the other side, but what's the alternative? To hold on to this thing that might be slowly dying? That's not safe either. So leaning into change feels to me like the most important thing for folks to take away.
I think as part of that, I would just say grief is part of this, and sometimes grief needs a funeral, or do what you got to do, you know, to be sad. Let your teams be sad. Let the people around you make space for that, because with change comes shedding things that you loved, and that sucks.
I think one of the other things that feels like it really came out, Lenny, is there is opportunity here. And we got to start ignoring or muting or whatever some of this fear narrative, because right now the data isn't saying there are no jobs for you. The data is saying actually there's going to be a lot of opportunity created by this, particularly if you're willing to lean into reinvention, right? Not what was the form of your job two years ago, but what might it be in five years. What else? I guess breaking down walls is the other thing I was thinking. But what came to your list?
Well, there's the Manoush kind of perspective of imagine your job isn't going to go away. Imagine your job will continue but be different.
Yeah, exactly. I love... it was a real mind change for me when she said that, because I was like, whoa, yeah, what is engineering going to look like in five years?
Right. It's not going to go away. It's going to be something different.
Maybe there'll be more of it. Like journalism, there's more journalists out there, for better or for worse.
Yeah. And so far there's more PMs, more product, more engineers. Yeah. So far that's going well. Another takeaway is just be deliberate about what you give to AI. And there's a few ways to think about it. One is just, what is AI going to do a terrible job at? Don't do that. What is it you love that you don't want it to take from you? How do you avoid brain rot and just over-relying on AI? There's a few things there. Just don't give everything to AI.
Yeah. And I think also, for a lot of us, acknowledging that AI is not this super intelligent being. It is much more like an intern. And that means literally think about it as an intern. What would you give to an intern, and what would you never give them?
I think that's the core of it.
Yeah. And I feel like we said this, but there just may be some things we should never give away, and that is genuinely different. Give away your Legos, but be discerning about what you should keep and what should be human and what humans are here for. I think that's part of the conversation for the next, you know, however many years, is what are humans uniquely good at?
And what I think about there, just very concretely in my mind, is be the person that guides what you're doing to the vision you have of the world you want, not just the simplest thing, not the thing that AI is kind of pointing to. Just, what do you actually want? And that's kind of our job for a while.
Yeah, and at the individual level, how can this new tool, this unintelligent intern that you've hired, let you do more of what you love? You know, let you bring more joy into the world. Let you lean into the things that you love doing. How can it amplify you, instead of feeling like all you're doing is losing things to it?
What a moment in time we're in, Molly, right now.
Trying to figure all this out.
We're going to rewatch this in a year or two and make fun of ourselves.
Yeah, we're going to be... We nailed it. We got everything right.
Perfect.
Visionaries.
Oh, man. Well, with that, Molly, is there anything else? Anything you want to plug? Anything you want to point people to? Anything you want to tell people about that you're up to that might be interesting or useful to them?
Well, I'll say two things. First, I get a lot of wisdom, but also a lot of support for this kind of conversation, from the community in Glue Club. And I think if you're a leader listening to this and feeling lonely and a little bit unsupported, find a community, whether it's Glue Club or another one, where you can talk about this, because I do think leaning on each other to learn what's real and what's not and what's valuable and what's not... so much of what we're doing in Glue Club is being like, is anybody else seeing this? And that is really valuable. You can do it in Lenny land as well, but just get out there and talk about what's real and what's not. Don't just stay in a sad, lonely hole.
But the second thing I wanted to say, Lenny, is thank you for making space for this. I think it's important. There's so much space right now for the what should I do, and what skills should I get, and what course should I take, and what tool should I try next, who's hot today. But we aren't talking enough about the human side of this and what it feels like to be going through this for everyone. And so I'm grateful that you reached out and we talked about making space to talk about this, because hopefully people will leave feeling a little more sane and like, oh, everybody else is feeling a little crazy right now, too. And there's lots of people out there grieving, and that's okay. That's normal. So, thank you, Lenny.
Thank you, Molly. I have a term for these sorts of episodes. I call them Trojan horse episodes, where people come for all this tactical, concrete stuff: how do I use AI in this way? How do I become a better PM? And then there's these—
You get therapy.
Important episodes that slip right in. Yeah, exactly. You're like, "Oh, okay. I didn't know I needed that, but that was very helpful."
Yeah.
So, well, welcome to the Trojan horse smuggling in.
I love it.
Very timely with the Odyssey these days, huh?
Yeah, dude.
All right, Molly. Well, thank you so much for being here.
Yeah. Thanks for having me, Lenny.
Bye, everyone.
Bye, everybody.
Thank you so much for listening. If you found this valuable, you can subscribe to the show on Apple Podcasts, Spotify, or your favorite podcast app. Also, please consider giving us a rating or leaving a review, as that really helps other listeners find the podcast. You can find all past episodes or learn more about the show at lennyspodcast.com. See you in the next episode.
Article published