Dan Shipper's Predictions for AI and Work: Super-Agents, Codex as the New Desktop, and Why He Would Buy SaaS Stocks
Lenny's Podcast
A year before this conversation, Dan Shipper, co-founder and CEO of Every, told Lenny Rachitsky that people were sleeping on Claude Code, especially for non-engineering work. Rachitsky says that call turned out to be remarkably accurate. Anthropic went on to build Claude Cowork on that idea, and OpenAI's Codex is now moving in the same direction. So Rachitsky invited Shipper back to make a new set of predictions, which the two agreed to score in May 2027.
The predictions fall into three groups: how the way we work will change, how the shape of the work itself will change, and who will do well. Throughout, Shipper describes himself as "simultaneously extremely AI-pilled and very bullish on humans." He expects AI to reshape nearly every workflow. He also argues that it will not eliminate human work, SaaS, or most jobs, and that it may create more work than it removes.
How Every "predicts" the future: living in it
Shipper says Every tries not to prognosticate. Instead, the team lives in the future together. The company has grown from about 15 people at his last appearance to almost 30. Every employee is an AI early adopter, including the engineers, designers, writers, editors, salespeople, and customer service staff. He contrasts this with most companies, where early adopters work alongside a middle-of-the-pack majority and a strongly anti-AI contingent. Every also reviews models, so the team often gets alpha and beta access to new releases and can help steer them a little.
In that setting, he says, prediction mostly means noticing what is happening and writing about it. Articulating what you notice makes it real for the team and for readers.
His original Claude Code call came about this way. The team tried Claude Code at launch as part of its normal job, and it was a little early. Around Sonnet 3.5 or 3.7, during their "vibe check," they realized Anthropic had effectively gotten rid of the code editor. From then on, Every shifted to a world where no one looked at code and everyone talked to their computer in English through Claude Code in the terminal. The company ran two or three internal software products at the time and runs about six now. Shipper then started wondering whether the tool could help with his writing. He uses an internal "reach test" for new tools: when you wake up in the morning, do you reach for it without thinking?
The paradox behind the predictions
Shipper opens with a tension. Benchmarks suggest AI will soon do everything. He cites METR's measure of how long models can work autonomously and says Anthropic's Mythos Preview can reportedly handle 17-hour tasks at 50% accuracy. He considers that progress real and exponential. Even so, his feeling is that a year from now "we actually have a lot more work to do. Humans have a lot more work to do," even as models get better at doing work.
He adds that his timelines are loose. Most of what he predicts should be apparent within a year, but it could take longer. To count as correct, a prediction should at least look like it is moving in the right direction by then.
Prediction 1: One super-agent per company, not a daemon for everyone
Shipper's main prediction about how work will change is that it will split into two modes. The first is the one people imagined years ago: everyone at a company will have at least one agent they talk to and offload work to, essentially something like OpenClaw.
He says the architecture of that setup is still unsettled. It could be one agent per person, per team, or per company, or some kind of parallel "shadow org chart" of specialized agents. When OpenClaw came out, everyone at Every adopted it, and Shipper was convinced each person would have their own agent. He compares a personal agent to the daemons in The Golden Compass, a small reflection of you that sits on your shoulder.
He has since "completely flipped." He now thinks the near-term model is a single super-agent for the whole company, and he points to Shopify and Ramp as companies that already have one. He still expects personal agents to arrive eventually. For now, though, the OpenClaw hype ran into reality at Every. Setting it up was too much work, it broke constantly, and it required things like SSH-ing into a server. Most people either could not or would not spend that time.
The underlying reason, in his view, applies to any harness. For an agent to be useful today, "it really needs a human who cares about it": someone who watches what it does, makes sure it is doing the right thing, and keeps it useful. Once that person stops caring and decides they no longer want to maintain the thing, the agent stops being useful.
That is why he sees companies consolidating around one agent, maintained by a forward deployed engineer or someone with a similar profile. The agent sometimes handles everything. Often it handles one job the company has decided everyone needs, such as data requests. He expects specialization to trickle down from there to team agents and eventually personal ones as people get used to the use cases and agents become "less fiddly." He says the model companies themselves are starting to see this pattern.
Asked where these agents will live, Shipper says work agents will mostly be in Slack. People enjoy using OpenClaw through iMessage on their phones, but he believes people want to keep personal and work agents separate. Every's COO, Brandon Gell, calls the personal side "computer errands," like ordering groceries. Shipper thinks that territory will be huge, but Every focuses on work.
Prediction 2: Codex and Cowork become the operating system for work
The second mode is that most of the work you do on your computer will happen inside an environment like Codex or Claude Cowork. That environment becomes the operating system for email, documents, and everything else. Shipper calls this the clear competitive landscape.
His account of how this happened: Anthropic realized that an agent running locally has access to everything the user has access to. It uses the terminal, which gives it powerful access, and models are very good at the terminal because so much material about it exists online. At a time when others still thought of coding models as pair programmers, Anthropic made the local agent work. Earlier cloud-environment approaches, he says, such as Devin and OpenAI's own attempt, did not see the same adoption. Then Anthropic and its community discovered that a local coding agent that can build anything is also good at any kind of work. People hacked Claude Code to do all their work, and Anthropic built Cowork as "a little bit of a nicer wrapping" around the same thing.
Shipper thinks OpenAI was very far behind for a while and has recently pulled ahead. In his description, earlier Codex versions were very smart but overly literal. They did exactly what you said rather than what you meant. Around the 5.3 release, perhaps three or four months before this conversation, OpenAI moved toward a fast model suited to general knowledge work and launched the Codex desktop app. He reads the app as OpenAI skipping straight to where Anthropic had arrived after going from Claude Code to Cowork, a path he says is visible in the tabs of Anthropic's desktop app. He calls the competition a horse race in which leadership will change. Right now, Codex is his daily driver.
The feature he considers most important is Codex's in-app browser. He keeps one Codex thread per project. To write, he opens the in-app browser to a document in Proof, a markdown editor he built, and lets Codex watch. He can see what Codex is doing and Codex can see what he is doing, all in one place. He describes this as an extension of what made Claude Code work in the first place. The result feels like a parallel work buddy that can edit the document, go do research, and use his computer for anything he could do.
Email is his favorite example. He says he has been at inbox zero for about ten days straight, which he calls unprecedented for him. Codex gathers his emails using Cora, Every's email agent, and renders a small page. He then monologues through the emails one by one. One instruction might be to research something. Another, responding to a question from the company's lawyers, might be to collect four years of documents into a report and send it. The tasks he used to procrastinate on now just get done.
He points out that this inverts the common assumption. People expected the ideal was to put AI into the browser. What he finds valuable instead is putting a browser into the agent you already use, so it can see everything you are doing. He says this combination is rare today, noting that Claude Code does not let you browse external websites inside it, and he predicts it will be common within a year.
What this means for SaaS: bring your own tokens, and build for humans and agents together
Rachitsky highlights one implication: SaaS tools will increasingly run inside Codex or Claude Code rather than embedding their own AI. Shipper agrees and describes the second-order effect. When he uses Proof, PostHog, or any website inside his agent, the agent has his access, and it runs on his tokens rather than the vendor's. That "puts SaaS back in its place." Vendors need to make their products agent-friendly. Everyone now has a CLI, the HTML should be usable, and anything that happens through the CLI should appear to the user immediately. What vendors don't necessarily need is their own built-in agent as the main interface. Proof, for example, pays nothing for tokens because users bring their own AI. Shipper argues this changes both what SaaS companies build and their margins.
He expects this to produce a new category of software. Classic productivity tools like Slack, Word, and PowerPoint were built for humans. CLIs are built for agents working independently. The new paradigm has a human and an agent working on the same piece of work at the same time, each seeing what the other is doing. That changes product design in several ways:
- Simpler products. Proof lacks much of Word's formatting, page breaks, and table machinery, because the agent handles formatting.
- New affordances. Agents can make huge numbers of changes at once. Products need approval flows, an inbox summarizing what will happen or has happened, logs, and fast rollback.
- Different infrastructure. Agents can make a very large number of requests in seconds. Shipper attributes GitHub's current problems to exploding usage that is largely agents.
He also gives a concrete benefit. Users of Proof and some other Every products don't email support. Their agents file bug reports that include exact repro steps and, because Proof is open source, a guess at where the problem is in the code. These become GitHub issues, and Every can send an agent to fix them. It doesn't work for everything, but he sees the beginnings of a fast closed loop: a user hits a paper cut, their agent talks to the company's agent, and the company's agent fixes it.
Where Cursor fits, and the race to own the harness
Asked about Cursor, Shipper says it sees many of the same things. In some respects it is ahead: he considers Cursor's cloud implementation better and more advanced than OpenAI's or Anthropic's. However, Cursor has chosen more clearly to serve programmers. The definition of "programmer" is expanding enough to give it a large market, but he doubts it will move into general work like making slide decks, and he thinks deciding whether to stay coders-only will be hard for the company.
More broadly, he says every model company now understands that a harness is needed to get the most out of a model. Platforms are moving from prompt-and-response APIs toward running the model on a cloud computer and returning the result. He cites Anthropic's managed agents offering, assumes OpenAI will respond, and describes Cursor as having been "essentially acquired by SpaceX," not a full acquisition but close. In his view, the ultimate form of this harness is one that can do any kind of knowledge work.
"CLIs are over"
Rachitsky asks whether this means people will move away from working in the terminal. Shipper says "we speed ran the CLI era." CLIs have existed for decades and will keep existing. But during Claude Code's rise, people concluded the CLI was the thing that worked, and Shipper thinks that was a mistake. Once you use a real UI for agents, you remember "we made GUIs for a reason." You get the same benefits in a GUI, especially for non-programming work. He estimates that most of Every's technical staff no longer use a CLI as their main work surface. Some programmers still drop into it occasionally, but mostly they work in Codex, Claude Code, or Cursor.
Two agents are better than one
Shipper says he is "definitely not an agent maximalist" and expects people to use many agents. He thinks two agents often beat one. When Codex talks to another agent, it can convey far more context about him than he would type himself.
His example is Every's hosted OpenClaw product. The company had it on a waitlist and had to pause after starting to let people in. OpenClaw moves so fast that a platform built on it can't fix things when they break. Along the way, the team learned something about onboarding. The conventional approach is to build a web form or Slack workflow asking who the user is, what they want, and what their dream outcome is. The alternative is to serve only users of Codex or Cowork. The user pastes a prompt into their agent, the agent talks to the app (a plain server or another agent), and because Codex already knows what Shipper has been working on and how he might use the app, the result is highly customized. When something breaks in a technical product, he can tell Codex to go fix it, and Codex talks to the app to work out the problem. He believes the whole paradigm changes once you assume everyone has an agent that talks to other agents.
Why he would buy SaaS stocks
Shipper's most contrarian call is that the "SaaS apocalypse" is wrong and that SaaS stocks will rise sharply over the next couple of years. He adds that this is not investment advice.
His reasoning ties back to tokens. Many SaaS companies are rushing to bolt on agents on the assumption that agents will become the main interface, and that costs tokens. If users treat Codex or Cowork as their main work surface and still want SaaS, the vendor's job becomes building software that humans and AI can collaborate on. He considers that hard to do, but much cheaper to run than paying for everyone's tokens.
He also points to Every's own experience. Everyone there has agents and uses Codex, yet the company still pays for a lot of SaaS, its SaaS spend is up year over year, and it does not vibe-code every small tool. His thesis is that agents increase the number of SaaS users, so SaaS companies will see a huge spike in demand from agents using their products at high volume. That creates infrastructure and pricing challenges, but it makes him bullish.
"Automation is a lie": the senior engineer benchmark
Rachitsky notes that Every doubled its headcount, which is not what one would expect from such an AI-forward company. Shipper's explanation is that "automation is a lie," in the sense that every automation needs a human watching it to make sure it works. He returns to an idea from an earlier essay of his, the "allocation economy," in which working with AI resembles being a manager. Managers, he points out, work a lot. They check in constantly and figure out how to make things better. Managing models differs from managing people in some ways, but it still takes time and attention. He says he feels this paradox personally: there is more automation than ever, and he works much more.
What clarified the paradox for him was building his own benchmark. He vibe-coded Proof on the side while running Every. It worked in internal testing and with beta testers, but after launch the servers went down roughly every ten minutes. He couldn't fix it. Codex either didn't know what was wrong or claimed to have fixed it while causing four new errors. He wasn't sleeping and says he vibe-coded so hard he got bursitis in his elbow. Eventually, two senior engineers independently rewrote the codebase.
Those two rewrites became his "senior engineer benchmark." Each new model gets a prompt saying, in effect, that the code is vibe-coded slop and it should rewrite it from first principles. By his account, every model before GPT-5.5 scored around 30 out of 100, while the human senior engineers scored in the high 80s to low 90s. GPT-5.5 scored about 62, and that result used a plan written by Opus 4.7. Shipper considers Opus 4.7's plans very good, but says GPT-5.5 is the only model with the agency and confidence to actually rip out old code and rewrite it. Other models paper over the edges with small patches even when told not to. From that jump, he finds it "very clear" the benchmark will reach senior-engineer level within a year or less.
His caveat is that he will easily be able to redesign the benchmark to zero out the model at that point. The current prompt looks effortless, but it took him a while to find wording that reveals the model's capability without giving away the answer. His original prompt, the one he actually used during the outage, asked the model to review the four or five issues reported the day before, make a plan to resolve them, and carry it out. Shipper predicts that every coding model on the market will still take that instruction at face value a year from now and simply try to fix the listed issues. A human senior engineer, by contrast, looks at the codebase, concludes it is badly built, and says a risky rewrite is needed even though you don't want to hear it. A model might get there if you asked directly, but it won't do so on its own, and there are incentives pushing against it. There is also always a higher frame a human can step up to.
His broader point is that benchmarks rise on problems that have already been framed, articulated, and made scorable. The thinking needed to frame a problem in the first place can't be measured. So saturating a benchmark doesn't mean replacing all senior engineers, and that is why he still hires them.
Rachitsky mentions hearing that data-labeling companies are buying pre-2021 human-written code. Shipper clarifies that his benchmark's rewrites were not hand-written. The engineers used AI. Comparing an unaided human with AI would be like racing a person against a car. The meaningful comparison is between one human using AI and another, "cuz AI doesn't use itself." The engineers simply used AI in a way he couldn't, because he didn't understand the codebase and didn't have time to learn it.
Shipper also adds a refinement to Rachitsky's takeaways for this section. The current approach to agent-friendly software is a CLI the agent uses on a task you delegated. Where things are heading, the agent probably uses the CLI while you use the web interface, and the two have to stay in sync.
The changing shape of work: floods of pull requests and forward deployed engineers
Once people have both delegated agents and a Codex-style work surface, Shipper says, the volume of pull requests skyrockets. Every sees this internally, and he says the big model companies see it too. Consultants, ops staff, and editors now make PRs. A larger share of a company, or of its users, does work that used to require technical skills. That puts pressure on the people who have to absorb all the new code.
He cites OpenClaw creator Pete, who he says receives thousands of pull requests a day, spins up large numbers of Codex instances to sort them, and merges around a thousand. Shipper expects this to become more common. It raises the question of which PRs to merge. Adding capacity in one part of a process breaks other parts. Building used to be the hard part. Now the harder questions are whether something fits with the rest of the product, how to keep a coherent whole, and what to delete. He credits Anthropic with deleting a lot from Claude Code to keep it from bloating.
Rachitsky raises the role confusion people report when engineers can design, PMs can code, and marketers can ship. Shipper agrees it is real. At Every, everyone is something of a generalist, which he thinks suits smaller companies especially well. He expects things to settle: marketers will still do marketing, and touching the website will simply be part of the job.
He sees one new role as clearly real: the forward deployed engineer, which follows from "every agent needs a human." Big model companies have teams running their internal agents, and he doesn't expect those teams to disappear. Agents will grow more capable and more numerous, but people will still manage them. Every has a few such people and also lends them out through its consulting practice. He notes the irony that automation seems to have created jobs rather than removed them. His example is Nitesh, an AI engineer who spends much of his time in Slack talking to Claudie, the internal agent that runs Every's consulting practice. The work includes code and Claude Code, but much of it is conversations along the lines of "Why did you do this dumb thing?" Shipper says certain engineers love this kind of hands-on work with the newest tools.
He resists Rachitsky's framing of this as "babysitting." He describes it as an engineering challenge: building a system that lets people with less knowledge do what used to be technical work without doing something dumb.
Data scientists and the flood of bad analysis
Rachitsky relays a data science friend's complaint. Everyone now runs their own analyses, and the team spends most of its time reviewing incorrect work. Shipper sees this as a sign that the right systems haven't been set up. He describes at least one major model company that has a data science bot anyone in the org can query. It is connected to the data warehouse and knows who has permission to see what, and a dedicated team keeps it answering basic questions correctly. Without that team, he says, the data scientists "would hate their lives." With it, their jobs can actually improve, because routine requests are filtered out and they can focus on deeper questions.
Which roles have changed least?
Rachitsky asks which product or tech role has changed least. Shipper's first answer is that CEOs and investors can still largely opt out, and Rachitsky agrees it seems that way. Shipper then argues the opposite should be true. From Every's work with senior leadership teams, he believes "your company's only going to go as far as your CEO goes in AI," and that this can't be delegated because leaders need hands-on intuition. For now, though, a CEO's day, and perhaps a middle manager's, can look much the same as before. He expects that to change quickly at some point.
His other candidate is sales, because selling is so personal. AI is creeping into BDR-type work and research, and it helps with top-of-funnel sourcing and qualification, but the core job hasn't fundamentally changed. Rachitsky contrasts this with customer support, which has. Shipper's research example comes from recruiting. Every was hiring a head of learning and development, and he guessed that a former General Assembly employee who had become deeply interested in AI would be ideal. He typed that into Codex and left. When he came back, it had found someone who had been an instructor at General Assembly, was very into AI, and already followed Shipper on Twitter. Shipper DM'd him and they had dinner.
We will read far more AI-generated writing, and like it
Shipper predicts we will read much more AI-generated writing in documents and email and will be glad of it. Coding already works this way: he wouldn't want an engineer to hand-write a plan document. For Every's end-of-2025 quarterly planning, the team used a Notion agent. Against a top-level company strategy, each person talked to the agent. It asked about the past year, their goals, and their metrics, pushed back, and connected their answers to company strategy. Shipper then received what he calls "incredibly good" AI-generated plans for each team. They made it easier to see which teams needed to talk to each other without realizing it, and which plans were weak.
His definition of slop has two parts. A slop document took less time to make than it takes to read, and the sender doesn't stand behind every line. Sending him an AI-generated document is fine. Being unable to discuss what's in it is not allowed. He argues that most people are bad at writing strategy documents, so a GPT-5.5 draft well directed by a team member beats one typed out by hand.
Most of his own email, he says, is now written by GPT-5.5 in Codex, and he would prefer it to say so and might configure it that way. Codex usually asks him before sending. Once it didn't, and it sent an email to an investor that he hadn't looked at. When he checked his sent folder, it was exactly what he would have written. It can be a bit over-formal. His view is that most email is rote: he wants to decide what it says, but the actual sentences usually don't matter much to him.
He stresses that he is a writer who cares a great deal about writing and considers human writing important. Every publishes a mix of human and AI writing and always labels it; sometimes an AI co-author is useful. Every's guides are often agent-assisted and written to be read by both humans and agents. Shipper's reasoning is that the best way to apply a long informational piece is to have your agent ingest it and bring it up at the right moment, say the next time you work on pricing. The human reads the story and core ideas, and the agent holds the details. Rachitsky notes that people usually associate AI writing with social media and video, and that internal documents are a much less fraught setting.
Who wins: AI-pilled PMs and full-stack designers
Shipper says he is "super, super bullish on PMs." His evidence is Marcus, who runs Spiral, Every's writing app. Marcus is a PM by training who previously ran Axios's writing product, leading a large team to tens of millions in revenue. After taking a year off and going deep on AI, first with Cursor and now, Shipper believes, Claude Code, Marcus became what Shipper calls "lightly technical." He knows what a database migration is and can follow code if needed. Shipper says Every could never have hired him for this role a year earlier. The coding models are now good enough that his limited technical knowledge, paired with sharp product sense, writing sense, and attention to users, makes him formidable. According to Shipper, he ships faster than almost anyone on the team and turns every user conversation into a story about what to build and fix next, without needing to organize a team to do it. Rachitsky, who has long argued that deciding what to build is the skill that remains once building is automated, jokes that this is the most contrarian episode they have done.
Full-stack designers are Shipper's other big winners. Designers have long watched engineers resist or mis-implement their interactions. Now they can build those interactions themselves, and Shipper sees them submitting pull requests instead of handing work off. That matters because distinctive design is exactly what's hard to get from vibe coding, where everything looks the same. He sees a large opportunity for designers to improve and even start their own companies. Rachitsky agrees that creativity will matter more as a way to stand out from the slop. They note that designer hiring, according to Rachitsky's job-market research, hasn't grown in a while, and they flag it as something to watch.
No job apocalypse, and how to "ride the models"
Shipper believes "the AI job apocalypse is not really a thing." Companies are reorganizing, and he suspects much of that is overhiring or weak performance with AI as the excuse. He doesn't expect the mass unemployment some AI CEOs describe. He admits he has no crystal ball but says he has seen enough model releases to form a view. Models "make yesterday's human competence cheap." They take what has already been done and make it cheap to deploy anywhere, so it spreads fast: landing pages everywhere, slop tweets everywhere. Because everyone uses roughly the same models, default output looks the same and becomes commoditized. Humans then take that "frozen human competence from yesterday" and make something new for their particular situation. He argues that the models' nature, combined with the incentive for model companies to make them compliant and aligned, means they will structurally trail the people pushing ahead. When those advances get absorbed into models, more room opens up. That's why, in his view, everyone becoming an engineer creates more demand for engineers, who are needed to decide how all that output should actually fit into a codebase.
Rachitsky worries this could sound too comforting and asks what people should actually do. Shipper's answer is that "the only thing you need to do is ride the models." Use them for whatever you do, work inside tools like Codex and Cowork, and try each new model to see what new powers it offers, rather than ignoring it out of fear, which he calls a rational response. For someone at a company like Salesforce, which may not give employees access to the latest models, he suggests doing it in their own time. His personal practice is to play. He keeps a set of tasks models can't yet do and "turns the rock over again" with each release, as with his senior engineer benchmark jumping to about 60.
He argues that the edge of AI isn't in San Francisco. It is wherever AI meets a real person doing something. The builders don't fully know how their models will be used, so anyone trying a new model can be among the first to discover what it's good for. He thinks Every, based in Brooklyn, is well ahead of many people in San Francisco because the team uses AI for everything. Rachitsky adds that nearly anyone can get access to frontier models on release day, though Shipper jokes that seeing Anthropic staff using Mythos at an event annoyed him. Shipper says that if IBM had invented AI, it would cost a fortune and be used in uninteresting ways. He considers it important that AI came out of a Silicon Valley culture that wants intelligence "too cheap to meter."
Rachitsky observes a two-sided picture: SaaS, jobs, email, and Slack persist while every role transforms. Shipper agrees and offers a metaphor. Our intuitions about the future resemble medieval ideas about the edge of the map, where either dragons or utopia wait. When you get there, you find some good things, some bad things, and another horizon. He warns against getting carried away with stories that sound real in the moment. He also calls AI companies' fear-based messaging ineffective and wrong, while granting that it's genuine.
Closing advice and lightning round
Shipper's closing advice is to ride the models. Try all your workflows in Codex or Cowork. If your company won't allow it, experiment on your own time with agent products like OpenClaw or Hermes, or simpler options for less technical people, so you get comfortable with both modes of work. Above all, he says, try to have fun. FOMO about losing your job is a worse driver of useful discoveries than doing something enjoyable. Rachitsky echoes a past guest's advice to find your "moment of joy" with AI by picking a real problem and trying to solve it with tools like Lovable, Claude Code, or Replit.
In the lightning round, Shipper recommended three books:
- The last chapter of Annie Dillard's The Writing Life, which every Every employee receives and is expected to read.
- Churchill's history of World War II, which he values as the account of someone who both did the work and wrote about it.
- The Rigor of Angels, which links Heisenberg, Borges, and Kant and which he found full of overlaps with AI.
He is also watching Knicks basketball, which he took up this year, and the docuseries The Dark Wizard about climber and BASE jumper Dean Potter.
He named two mottos. The first is one he wrote in college: "Do things worth writing about and write things worth reading." The second comes from a meditation teacher he admires, who has since died: meet hard things from "a position of spaciousness and strength." He says this has helped him face fears about how AI might change his job.
Asked for an underrated AI tool, he said Codex again. He says he recently urged Boris and Kat from the Claude Code team at an Anthropic conference to try it, especially for tasks like email and analytics. He thinks Anthropic can catch up. He sees it as a horse race in which OpenAI has "gotten back the mandate of heaven a little bit" after a rough six months or so. He says he isn't sponsored by OpenAI, still uses Claude a lot, and would switch if the lead changed, just as he was loud about Claude Code when it was his favorite. Whether his predictions hold up is what he and Rachitsky plan to check in May 2027.
The last time you were on this podcast, you had this hot take that people were sleeping on Claude Code. You were so unbelievably right. The premise of this episode is we're going to go through what else you predict will happen.
The AI job apocalypse is not really a thing. I am super, super bullish on PMs and full-stack designers.
So, you guys are hiring doubled in people in the past year, which is not what people would have expected from a company that is so AI forward.
I'm simultaneously extremely AI-pilled and very bullish on humans. Automation is a lie. Every agent needs a human. We have so much automation, so much AI, and I also work way more.
Creativity. It just feels like it's going to be more and more valuable to stand out from all the slop that people are shipping and launching constantly.
What models do in general is they make yesterday's human competence cheap. And so, it becomes commoditized. It's not valuable anymore. What humans do is we go in there and we're like, "Yeah, we have all this frozen human competence from yesterday. How do I use this to make something new and interesting?"
What are some predictions for how the way we work is going to change?
It's going to bifurcate in two main ways. One is everyone's going to have at least one agent that they talk to, that they can offload work to. Second is that most of the work that you do is actually going to happen on your computer in an environment like Codex or Claude Cowork.
What you're predicting here is the SaaS tools will run within Codex or Claude Code.
I think the SaaS apocalypse is dumb. I would buy SaaS stocks right now. What agents do is increase the number of users of SaaS, not get rid of it.
A lot of people are moving to CLI and trying to work from the terminal.
We speedran the CLI era. It was nice while it lasted, but I think CLIs are over.
Today my guest is Dan Shipper, CEO and founder of Every. Dan and his team are building maybe the most AI forward startup out there. And as a result, are very much living in the future of how work is going to look as AI becomes a bigger and bigger part of our day-to-day. Everybody at their company, including every non-technical person, uses Codex and Cowork and Claude Code to get much of their work done. And this is why, way before anybody else, Dan saw the rise of Claude Code and what is now Cowork, which he predicted almost a year ago when he was on the podcast last time. So, I asked Dan to come back on the podcast to share his current biggest predictions for how work is going to change over the coming year for most people. We chatted about what work will look like at most companies at the end of this year, how the shape of the work we do will change, and who will do best in this coming future / what you need to be working on right now. Hint hint, product managers and designers are going to do very well. Dan makes a lot of bold predictions and many quite contrarian takes that I was not expecting him to say, and we are going to revisit this conversation exactly a year from today to see how much he got right.
Before we get into it, do not forget to check out Lenny's Product Hunt dot com for a free year of the hottest and most well-crafted AI products in the world available exclusively to Lenny's newsletter subscribers. With that, I bring you Dan Shipper.
Dan, thank you so much for being here and welcome back to the podcast.
Thanks for having me. Always a pleasure to be with you.
The last time you were on this podcast, you had this kind of, it was almost like an offhand hot take that people were sleeping on Claude Code, and in particular Claude Code for non-engineering work, for just fixing files, sorting your hard drive, just all these things that people hadn't thought about. Nobody was talking about this. This was a year ago.
You were so unbelievably right about this. It's just unreal what has happened since then. They built Cowork, which was this whole... They built on this very specific idea using Claude Code for non-technical work. And Codex is getting into this now. I imagine you've been seeing this. They're leaning into this non-technical use of basically coding agents. I feel like this has also been a big part of Anthropic's success over the past year, just how do non-technical people use this stuff?
So, you were just so go on this stuff. I even wrote a newsletter post building on this idea. I'm like, "Hey, this is interesting. I should dig into this." I asked people, "How do you use Claude Code for non-engineering work?" And I just had so many examples and it's my second most popular post.
So, clearly you have a unique glimpse into where things are heading. So, the premise of this episode is we're going to go through what else you predict will happen in the future, how things will change for people building products. And I think it'd be helpful to start with giving people a brief glimpse into just how you operate and how your team operates that gives you this unique lens into where things are going. So, just give us a sense of how you work.
Thank you. I really appreciate the introduction. And yeah, I think one of the things about predicting the future, or the way that we think about predicting the future at Every, is that what you don't want to do is prognosticate. What you want to do instead is just live in it together.
So, everybody at Every is an AI early adopter. We're almost 30 people now. I think when we did our interview we were 15, so we've doubled in size in the last year. We're all early adopters and we have engineers, we have designers, we have writers, we have editors, we have salespeople, we have customer service people. And everybody has a little bit of that, whatever that thing is where you're just like, "Oh, I like to explore. I like to experiment. I'm very curious and I'm super all in on AI." And what that does, I think, is it creates this little pocket of the future where we're all living in it together and we get to be a little bit further ahead, because at any other company there's a mix of people. There's early adopters, there's sort of the middle of the pack people, and there's people who are very anti.
And another thing that happens, which is really cool, is because of our role, you know, reviewing models and being a little bit of a tastemaker in AI, we get access to stuff before it comes out. So, we get to beta test and alpha test and kind of help steer the direction of where things are going a little bit, which is very, very cool.
And so, when I think about predicting the future, when you create an environment like that, it's actually just about noticing what's going on. And I think a core part of it, too, is writing about it. I think articulating what you're noticing, articulating the future, kind of brings it about in this way that makes it real for you and your team and then anybody else who's on the internet who's reading it.
And so, the Claude Code thing, this is this very organic thing where for us, we tried Claude Code when it came out. That's sort of our job. We try all the new stuff from the model companies. And at the time it was a little bit early. But right around, I think, Sonnet 3.5 or Sonnet 3.7, we were testing that to do our vibe check on it. And we're like, "Holy [ __ ] This is crazy. This is really... They got rid of the code editor." And so, from that point on, we just basically... At this point now we run like six software products internally. At that time we ran maybe two or three. And from that point on, we just started shifting to a world where no one was looking at the code. Everybody was, you know, talking to their computer in English using Claude Code in the terminal. And so, I was able to see, "Ooh, this is starting to happen."
And then because my job is a little bit to just push and play with stuff, I was like, "I wonder if I could use this for my writing. How could I do that?" And then it just starts to unfold and you're like, "Okay, this is not ready yet, but it's obviously useful for me." One of the things that we talk about internally is what I call the reach test, which is, when you wake up in the morning, do you reach for it organically?
I love this combination of you're using the latest stuff, and I think this is, as you said, maybe an underrated skill, you're good at being self-aware of here's what's weird and new and different and interesting. So, that's a really cool combination, partly because you have to write about it, and you write about it. So, I think that's the perfect recipe for someone having a sense of where things are going.
This episode is brought to you by our season's presenting sponsor WorkOS. What do OpenAI, Anthropic, Cursor, Vercel, 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.
So, the way that I'm going to structure this conversation, there's going to be basically three buckets of predictions. One is how the way we work is going to change in the coming years. Two is what the shape of the work we're going to be doing is going to look like and how it will change. And then three is who is going to be most successful in this future / what should you be doing and working on now to be successful in this future?
Lenny, my only ask is we come on a year from now and then you score it. I want to score it.
Okay, so this is a year from now. Okay. Okay. So, let's actually... Is this your predictions for in a year this is what it's going to look like, or this is the emerging future?
I will probably say I don't have an exact timeline. I think most of the stuff that I'm going to talk about will be pretty apparent within a year, but it may take longer than that. But I think it should within at least a year be not obviously wrong. It should seem like it's moving in that direction to count.
Okay. May of 2027, we will review your predictions.
Right. Amazing.
Confirmed. Okay, I love this. Okay, so let's dive in. What are some predictions for how the way we work is going to change in the coming year?
One of my favorite questions, because I think if you look at the benchmarks, you're just looking at, okay, yeah, AI is going to just take all of our jobs basically, you know.
METR has this really cool benchmark where it measures how long the newest models can do tasks autonomously, and it's like, oh, what's it called? Mythos Preview, the big Anthropic model that everyone's so worried about, it can do tasks of 17 hours at 50% accuracy. It's like, holy [ __ ] that's crazy. And I think it is real. It's true, and the model progress is going up exponentially.
And my experience and my feeling is that we will look back in a year and say we actually have a lot more work to do. Humans have a lot more work to do. Even as models get better at doing work. And there's a really interesting paradox there.
And my prediction for how work will... Well, my big prediction of how work will change, or how you will be doing work in a year, is it's going to bifurcate in two main ways, how you use agents.
One is you're going to be doing, I think, what we figured you would be doing 5 years ago when we thought about how work with AI works, which is everyone's going to have, at least in their company, at least one agent that they talk to that can do work, that they can offload work to. And we'll talk about what that looks like, but it's essentially like OpenClaw.
The second is that most of the work that you do is actually going to happen on your computer in an environment like Codex or Claude Cowork that becomes the sort of operating system for how you do all of your work, whether that's your email, the documents you create, all that kind of stuff. It's going to be on that kind of a surface. That's becoming the clear competitive landscape.
So I want to go in order of those two. So the first one is you're going to have agents you delegate to, probably in Slack, but, you know, anywhere. First thing that's interesting about that one is it's not clear what the architecture is going to be for that. Is everyone going to have an agent? Is every team going to have an agent? Is it going to be just one agent? Is it like agents specialize? There's this parallel shadow org chart.
And when OpenClaw first came out, everyone internally at Every adopted it, and I was very convinced that everyone would have their own agent. And there's some really interesting things about that world of, you know, a parallel org chart. Agents in that world sort of become little reflections of you, which is really cool and really interesting. Did you ever read The Golden Compass? It's like having a little daemon on your shoulder. You know, it's a little part of your soul. I really think that's sort of what it looked like was happening.
And so I was very into personal agents. And I have completely flipped. And I really think that the model for now is going to be a super agent, one agent for the entire company. And you're starting to see this in some companies. So Shopify very famously has one. Ramp has one now. And I think there's some really interesting reasons for that. I actually still think that the personal agent thing is coming.
But what we found is there's all this hype with OpenClaw. Everyone's like, "I'm going to set it up. It's so cool," or whatever. And then everyone realizes it's way too much work. This thing breaks all the time. I got to fumble around with it. I got to be able to SSH into my server and blah blah blah. And most people, to do work at least, just don't want to spend that time or can't.
And the fundamental underlying thing that drives that is, whether it's OpenClaw or any other harness, in order for an AI agent to be useful right now, it really needs a human who cares about it. It really needs a human personal connection with someone who's watching what it does and making sure that it's doing the right thing and that it's useful for people. And the minute you sever that connection, so the minute someone's like, "Ah, I don't want to maintain this dumb OpenClaw," is the minute the agent is not really that useful anymore. And that's why I think it has started to shift to a more one agent per company model, because for now the ideal is you basically set up a forward deployed engineer or someone with that sort of profile who's responsible for making
sure that that agent is working for the whole company. And then maybe you have some little team agents. And I think as the models get better at being more independent, that will shift down and it'll be more likely that we'll have more personal agents cuz we don't have to [ __ ] around with all the internals.
But the model that I see working for us and for a lot of other companies, including the model companies, the model companies themselves are starting to see this, is when it comes to the sort of async agents, you have one agent at the top that's doing sometimes it's everything, a lot of times it's a particular kind of job that you've decided that everyone in the company needs an agent for, like data requests. And then I think it will start at the top and then it sort of starts to trickle down where you make more specialized agents and teams and all that kind of stuff. And the mechanism is agents need people who care about them.
That is so interesting, that point about you need to garden your agent because there's context you have to keep adding to it. It breaks, as you said, and once it's just too much work, you're like, okay, forget this thing. I'm going to go back to Codex or Claude or something like that.
Exactly.
Okay, cool. So what you're predicting here is companies will have this super agent that everyone can talk to. As you said, Shopify's got River, I think it's called. What's the Ramp one called? I can't remember. Okay. It's probably got a fun name. Okay. So that's the prediction. That's the first prediction.
That's the first prediction. We will start with agents at the top that are more general and are used by more people in the company, and then it will start to kind of grow down as people get more used to these use cases, they get more specialized and agents become less fiddly. Like they just work better.
And is this mostly going to be in Slack, do you predict? For work?
Yeah, it seems to make sense. I think people love having the green bubbles on OpenClaw. I'm sorry, the blue bubbles on OpenClaw. Like if you can use it with your iPhone. But I think there's this little thing in people's heads where they really like to keep their personal and work agents separate.
And I think there's a whole territory. Our COO Brandon Gell calls this computer errands. There's this whole territory of using personal agents for your computer errands. It's like order my groceries or whatever, and there's so much of that that I think it's going to be huge for. But we focus mostly on the work stuff, and I think that's going to happen mostly in Slack.
Sweet. Go Slack. Do you want to talk about the other work surface?
Absolutely. Codex, Cowork. Okay. This is the last. Let's do it. I'm so excited about this one. I think it's the coolest thing.
So, basically what happened was Anthropic realized at some point that with Claude Code, if you put an agent on your computer and it runs on your computer, it has access to everything that you have access to. It uses the terminal, so it has basically superpowered access to it. And not only that, these agents really understand how to use the terminal cuz there's so much content online about that. And it created this superpowerful coding paradigm, which Anthropic was really doing first. OpenAI for a while was, in my opinion, very, very behind on this, and then in my opinion has surpassed them recently. It's really interesting.
But they were very early on this. When people were still thinking about coding agents or coding models as being really pair programmers, they were among the first to be like, "No." And do it successfully. There were people before them like Devin, who I think had the big cloud environment, and OpenAI tried this too, but the real adoption seems to have happened when you put it on your computer.
So, they figured that out. And then I think they figured out, along with their community, that once you have a coding agent on your computer that can build anything, it's actually really good for any kind of work you want to do. And people started just hacking Claude Code essentially to do all of their work. So, Anthropic then built Cowork, which is a little bit of a nicer wrapping around Claude Code, but it's fundamentally the same thing.
And then I think OpenAI made a couple of different bets, but their main bet on a programming agent was the earlier versions of Codex were very technical, and they were super smart, but they were a little bit autistic. It was a little hard to... they didn't quite get what you meant. They got exactly what you said.
And I think maybe three or four months ago, around the time that they launched 5.3, they started to move in this direction of, "Oh, no, we get it. This model is fast. It's really good for general purpose knowledge work type tasks." And then they launched the Codex desktop app. And if you look at all the lessons that Anthropic learned, they went from Claude Code to Cowork, and you can kind of see that in the tabs on the Anthropic desktop app UI. I think OpenAI was just like, "We see where this is going. Let's just skip to that."
And so, this is a horse race. They're going to have different positions. But I think Codex right now is my daily driver. I spend all my time in it basically. I flip to Claude every once in a while, but I think they're getting the paradigm right. And it's clear to me that whoever is in the lead, cuz again I think it'll change, it feels very obvious to me that all of the work that you do is going to be in one of those surfaces.
For example, when I'm writing a document, Codex has an in-app browser. And when I'm writing a document, I just go into one of my Codex threads, and I have one thread for every project. And I just open the in-app browser. I go to the document. I usually do it in Proof, which is this online markdown editor that I built. And then I just have Codex running and watching me in Proof. And Codex can see what I'm doing. I can see what Codex is doing. It's all kind of in one place, which is an extension of the same thing that made Claude Code work really well originally.
And I basically feel like I have this parallel work buddy that not only can respond and write in the document, but then it can go do research. It can use my computer to basically do anything that I can do on my computer. And that's incredibly powerful.
And I do this with everything. I've been at inbox zero for like 10 days straight now, which, if you know me, is crazy. I'm never like this. And that's because I literally just have Codex gather all my emails with Cora, which is our email agent. And then it renders a little page, and I think I showed you this at the Anthropic event. It renders a little page and I just monologue into it and just talk at each email. I'm like, "Okay, go research this. Oh, here's a question from our lawyers. Can you go collect all of the documents from the last four years and then put them into a report and send them?" And it just does it. And so all the stuff that I would procrastinate on, I don't really procrastinate on anymore.
And so I feel like for a long time, we thought, I thought too, that the optimal experience of AI was going to be take AI and put it in a browser. And I think the reverse is actually starting to happen and be really, really valuable in a way that I did not expect, which is take the AI agent that you use all the time on your computer and put a browser in it so it can see everything you're doing. And that is just a magical combination that is very uncommon now. You can't even do this in Claude Code, because they don't let you browse external websites inside of Claude Code. So it's very uncommon now, but I think it will be super common in a year.
This is more profound than it may even sound. What I'm hearing is, instead of AI being baked into SaaS tools, what you're predicting here is the SaaS tools will run within Codex or Claude Code.
That is one really important second-order effect of this. Okay, so yeah, I'm using Proof, or really any website, maybe PostHog or whatever. And I'm doing it inside of my agent. And the agent has access to the website, so it has access to everything that I have access to. And it has access to my whole computer. When I run the agent on that website, I'm using my tokens. I'm not using the vendor's tokens. I'm not using the app's tokens.
And so it puts SaaS back in its place where, yeah, you want to make it friendly for an agent. And everyone's got a CLI now. You want to make the HTML really usable. You want to make sure that anything that happens in the CLI shows up for the user immediately. All that kind of stuff. There are a lot of issues to deal with. But once you do that, you actually don't really need to think about having an AI surface that's primarily going to be the thing that users use, in the sense that you don't need to build an agent natively into your product.
I think you can, and there's another really interesting bifurcation of this that we should talk about, which is that having two agents is better than one. But I think for now there's this really cool thing where, with Proof for example, anyone who uses it, I don't pay for tokens because they bring their AI to Proof. And so it changes what you build as a SaaS company, and you build it now for both humans and agents to use at the same time, and it changes your margins back to, well, I don't really have to pay for tokens anymore cuz the user is going to bring AI. So I think this is a huge deal.
So what you're describing here is more and more work that we do, more and more professional work, is just going to happen within Codex or Claude Code. Where does Cursor fit into this? Is there potential there?
That's a good question. I think that Cursor sees a lot of the same stuff. And in some ways they have some of the same stuff but it's better. I think that Cursor's cloud implementation is better than either OpenAI's or Anthropic's and is more advanced. And I think that Cursor has, at least so far, more distinctly chosen a lane. They're more distinctly choosing to be for programmers. And that may limit how far they get in here. I think the definition of programmer is expanding enough that they'll have a big market, but I don't know that they're going to jump into, okay, use this to make a slide deck or whatever.
But it is really clear that every model company is starting to realize how important it is to have a harness to get the most out of the model. And so where all the platforms are moving is to a world where you're not just doing prompt and response when you call the model on the OpenAI platform or the Anthropic platform. They are literally running the model on a computer in the cloud that they run and then giving you the result out of it. And they know that in order to get the best results from the model they need to offer that. And so you see Anthropic's got Claude Managed Agents. OpenAI does not have a response yet, but I assume that that's going to happen. And now Cursor was just essentially acquired by SpaceX. It's not a full acquisition, but it's close.
So I think people are starting to realize, I can't just do the model part of it. I have to have this harness above it. And I think the ultimate form of that harness is, I can do any kind of knowledge work. Cursor itself, it feels like it's going to be a hard decision for them whether to stay just for coders or not.
So for people building products that aren't OpenAI or Anthropic, if this proves to be true, the prediction here is their users are going to be using their product over time inside of one of these agents. Is there something you would do if you were one of those companies to prepare for that future?
I would just prepare for that. So for example, every more classic piece of productivity software, whether it's Slack or Word docs or PowerPoints or whatever, is really mostly meant for a human to use. And now people are doing CLIs, so it's meant for an agent to use independently of a human. And we're moving into this new paradigm, I think, where the human and the agent are on the same piece of work together and they're both doing things. I need to have visibility into what the agent is doing. The agent has to have visibility into what I'm doing. We have to go back and forth in this sort of seamless way. And the kind of software that you make for that is going to be very different.
So, for example, there's a lot of stuff that Proof doesn't have. I don't have to have a lot of the Word document kind of formatting or page breaks or making tables or whatever, cuz the agent just does it. I don't need to worry about that. It can do all the formatting for me. So, you can make the products a lot simpler and faster to start than the legacy products are.
And then there's all these other affordances that you need to start to have because the way agents interact with software is very different. For example, agents can do a lot at once. They can just do a billion different things to your document or your slide deck or your code base or whatever. And how you display that to the user is going to be very different than the way you might display a human being concurrent in your document and doing stuff. You need approval. You need a sort of inbox that summarizes, here's all the stuff that's going to happen or has happened. You need logs and the ability to roll it back real quick.
So, there's all those kinds of considerations that change the actual product. And then the underlying UX of it or the underlying infrastructure you need is different too, because agents can make a billion requests in like 3 seconds. So, how are you going to deal with that, right? This is exactly why GitHub is having problems right now, because the number of people using GitHub is skyrocketing exponentially, and it's really just people's agents in GitHub.
So, I think it's this whole new world that you're just starting to see a little peek of. But there's so many cool things about it. So, for example, in Proof and some of our other products too, when someone has a problem, they don't email support. Their agent sends a bug report. And an agent bug report is way better than a human bug report. It has, here's exactly what I did, here's the exact repro steps, and Proof is open source, so here's what I think is going on in the code base. And then we just get that, it becomes a GitHub issue, and then we can just send off an agent to fix it. You can't do that with everything, but it's so much better.
And you can see the glimmers of this very fast closed loop between, I ran into something, a paper cut, a little feature I want, a little bug, and my agent just goes off and talks to the company agent, and then the company agent just goes and fixes it. That I think is incredibly cool.
So, is there a part of this that you... a lot of people are moving to a CLI and trying to work from the terminal. Is
part of this prediction that people shift away from that and back to actually GUIs with agents kind of running alongside them?
CLIs are over. We speed ran the CLI era. It was nice while it lasted, but I think it's pretty clear... It's not that CLIs are going to completely go away. Obviously, they've been around for the last like 30 years or 40 years or 50 years or whatever. They will continue to be around. And I think there was this moment when Claude Code was like so popular, or when Claude Code was really starting to gain in popularity, that people were like, the thing that's working is the fact that it's a CLI, and I don't think that's what it is.
And when you move into an actual UI for this, you start to realize we made GUIs for a reason. And it's just nicer to be in a GUI. And you can get all the same benefits inside of a GUI, especially for non-programmer work, but I would estimate that definitely the majority of the technical people inside of Every are not using CLIs anymore as their main work surface. I think a lot of programmers are still flipping into it every once in a while, but more or less they're using Codex, Claude Code, Cursor, that kind of thing.
Awesome. Okay. I definitely wanted to make that part clear. So, coming back to the big picture of the prediction here, there's kind of these two modes of work that you're anticipating. One is this kind of super agent within a company that you chat with through Slack most likely that can go off and do work and answer questions. And then there's, on your computer, running Codex or Claude Code. And within that, all the work that you normally do kind of on your computer is now going to be living within Codex or Claude Code or maybe some third party that emerges that we're not even aware of yet.
Yes, and you're going to use apps inside of the internal browser of those tools.
Wow, okay. Listening to you talk about it, it may not feel as profound as it is, 'cause this is a big change to how we work. We don't currently have an AI that we talk to regularly in Slack and we also don't work currently mostly in Codex or Claude Code. So, this is actually a pretty massive shift.
I think so.
Is there anything else along these lines before we get into our next prediction?
Well, a few things. I'm definitely not an agent maximalist. I really think we're going to have a lot of different agents that we use. Seems pretty clear to me. And I really do think that two agents are better than one. So, what's a good example? When I have Codex interact with another agent, it can give so much more context about me and what I want than I would be able to type. And it can go back and forth talking about things that would take a long time for me to express directly to an agent, that you get this like speed-up effect when you assume that your users are using Codex or Claude Code or Cowork as their basic way they access your app.
And a really simple example, we have this hosted OpenClaw product which we had on waitlist. We actually had to pause it because we started taking people off the waitlist, and OpenClaw is just a very hard agent harness to make work. It's moving so incredibly fast. And if you're a platform for it, when things break you can't fix it. It's very hard. But one of the things that we learned in that process is, if you're, let's say, building an agent product or any new software experience, what you would assume, let's say to set up an agent, is you need to build like a little web interface or a little Slack workflow that asks people, "Okay, who are you and what are you going to use this for and what's your ideal, you know, dream outcome?" Or whatever the things are that you would put on an onboarding checklist.
If instead you just make a hard line of, we are only going to service users who use Codex or Cowork, what happens is you just paste a prompt into Codex or Cowork. It goes and talks to the app, and the app can be either just a regular server or it can be its own agent. And Codex has so much information about you that it can just give it, here's all the stuff I've been working on with Dan, here's all the ways that he might want to use this app, and then bring it back to me, and it's this very custom experience. And also for a technical product like an agent, when something goes wrong, I can just tell Codex, "Go fix it." And Codex will go talk to the app and figure out what's going on for me. And so I think the whole paradigm starts to change when you assume that everyone's got an agent and those agents are talking to other agents in this really magical and important way.
There's a couple more things I want to touch on before we get started 'cause there's so much to talk about. One, you made this point about SaaS tools, like you can use tokens from the model companies basically when using a SaaS tool. Talk a bit more about that 'cause that may change the business model for SaaS companies in the future. That feels like a big deal.
Well, I think it actually may save their margins. Because right now all these companies are rushing to add an agent to their offering and thinking, "Oh, the agent is going to be the main way that people interact with me." And that costs tokens, obviously. And I actually think once I have Codex or Cowork as my main work surface, I still want to use SaaS. So, this is another good prediction. I would buy SaaS stocks right now. I think the SaaS apocalypse is dumb, and SaaS stocks will be up majorly in the next couple years. Not investment advice, but you know, I would buy SaaS stocks.
So, I think it saves your margin because now the way that you're thinking is not, I have to build AI into this. It's more like, I have to make a piece of software that humans and AI want to collaborate on together. And that's hard, but once you build it, it's a lot cheaper than assuming everyone's spending tokens. And I think it's a good business. And part of the reason I'm so bullish on SaaS is, A, everybody internally here, like I said, we've all got agents and we're all using Codex and whatever, and we still pay for a ton of SaaS, and our SaaS spend is up year over year. And we're not vibe coding every single little thing, you know? And I think that what agents do is increase the number of users of SaaS, not get rid of it. And so I think SaaS companies are going to see an insane spike in the amount of demand that they have because there's going to be tons of agents using these products at a very high volume. And like I said, that's a huge infrastructure challenge. There's a lot of interesting pricing challenges, but it makes me very bullish on SaaS.
I love that. If anything else comes out of this conversation: Dan Shipper, SaaS is the future of AI. Yeah, this is quite contrarian, and the other interesting piece is the fact that you guys are hiring, that you doubled in people in the past year, which is not what people would have expected from a company that is so AI forward. Talk about your experience there of just, okay, we still actually need humans.
Automation is a lie, in the sense that every time you automate something, in order to make sure the automation is working well, you need a human on top of it making sure that it's working well. And so, you know, I wrote this piece a couple years ago about the allocation economy, the idea that the way that humans are going to work with AI is going to be like being a manager. And the thing that you have to remember about managers is managers actually spend a lot of time working. Most managers are not on the beach. They're checking in with their employees all the time and trying to figure out, okay, how do we make this work well? How do we make it better? How's it doing? How's this person doing? All that kind of stuff. And I think there's some differences between being a human manager and being a model manager, but fundamentally it still requires a lot of time and attention.
And I think that we kind of miss that in the model discourse. And one of the reasons is benchmarks make it look like AI is more autonomous than it is. And by autonomy, I mean something specific, and I'm going to try to express this. It's a little hard to express, but I learned this for myself because I've been feeling this paradox a little bit. I've been feeling like we have so much automation, so much AI, and I also work way more. And I think part of the paradox started to resolve for me a little bit when I made my own benchmark.
So, I made this... It's called the senior engineer benchmark, and it's like, "How good is AI versus a human engineer?" And the way that I built it is, again, I have this app, Proof. I just vibe coded it on the side while running the rest of Every. And when we launched it, because it was completely vibe coded, it just started going down and I couldn't fix it. And it was very embarrassing. I had a lot of egg on my face. And the product worked. We tested it internally. We had a lot of beta testers, but the day after launch, every like 10 minutes the servers would go down, and people were looking at me and I'd be like, "I don't know what's going on." Like, "Codex, fix it." And Codex was like, "I don't know what's going on." Or really Codex was like, "I do know what's going on. I fixed it." And then it would cause four other errors, and then you're just going around in a circle, and I wasn't sleeping, and I vibe coded so hard I got bursitis on my elbow.
So, there's a life lesson in there. Vibe coder elbow.
So, anyway, I actually got two different senior engineers to fix it independently. So, I have two different rewrites of the code base that tell me how they did it, right? And so, what I get to do is, when we get new models, I just give the new model a prompt. I say, "This is vibe coded slop. If you wanted to rewrite it from first principles, how would you write it? Go do it."
And all the models until GPT-5.5 got like a 30 out of 100. And a human senior engineer gets like high 80s, low 90s out of 100. So, there's a lot to go. And then I tried GPT-5.5 and it got like a 62. And mind you, the 60 score was GPT-5.5 using an Opus 4.7 plan. Opus 4.7 plans are very good. GPT-5.5 is the only model though that has the sense of agency and confidence to just rip out old code and actually rewrite from first principles. Other coding models, they kind of end up papering over the edges, and they're like, "Oh, this is a big job. I'll just do a little patch." And you're like, "No, I specifically told you not to."
So, GPT-5.5, there's like a 30-point bump in the score. 60 out of 100. It's very clear that in a year or less, it's going to be senior engineer level, right? And that gives you a certain picture in your mind, especially based on how I named the benchmark, which I think a lot of benchmarks do. And I can tell you that when we get to that point, it will be very easy for me to change the benchmark to zero out the current model. So, that gets a zero out of 100.
And so, for example, it seems like there's no skill or no thought in the prompt, which is, this is vibe coded slop, fix it from first principles, but actually it took me a while to get to a prompt that didn't give away the answer but got the model to reveal what it's capable of. And the original prompt I gave it was the original prompt that I gave it when I was trying to fix the issue in production when it was going down, which is, I'd woken up in the morning and I was like, "Okay, we had four or five reported issues yesterday. I want you to go through all the issues and then make a plan for how to resolve all of them and go do it," right?
And every coding model on the market, and I'm pretty sure this... Here's a prediction. I'm pretty sure every coding model on the market will still do this in a year. Every coding model on the market will take that instruction seriously. And if I tell it, "Here's a bunch of issues, go fix it," they will just go try to fix the issues. What an actual human senior engineer does is they go look at the code base and they're like, "This is a piece of [ __ ]. This guy doesn't know what he's doing." And then they say, "We're going to have to actually rewrite a lot of this, and it's going to be hard and risky. I know you don't want to hear that, but we're going to have to do that." And if you asked the model, "Hey, should we do that?" it'll probably get there, but it's not going to do it on its own. And there's a lot of incentives pushing against it doing that. And even if it does that, there's always a higher frame for us to go.
And so I think it's really important, when we think about benchmark progress, to think about it from that perspective, which is benchmarks rise on problems that we've framed, that we can articulate, that we can score. And there's a lot of work that's human work that can't be scored until you write it down, but the act of thinking to prompt it or write it down is something that you can't measure, but kind of means that even if the benchmarks get saturated, it doesn't mean the same thing as you totally replace all senior engineers. And I think it's why, even though the models are getting better at automation, I still hire engineers.
I am so excited to tell you about this season's supporting sponsor, Vanta. Vanta helps over 15,000 companies like Cursor, Ramp, Duolingo, Snowflake, and Atlassian earn and prove trust with their customers. Teams are building and shipping products faster than ever thanks to AI. But as a result, the amount of risk being introduced into your product and your business is higher than it's ever been. Every security leader that I talk to is feeling the increasing weight of protecting their organization, their business, and not to mention their customer data. Because things are moving so fast, they are constantly reacting, having to guess at priorities, and having to make do with outdated solutions. Vanta automates compliance and risk management with over 35 security and privacy frameworks, including SOC 2, ISO 27001, and HIPAA. This helps companies get compliant fast and stay compliant. More than ever before, trust has the power to make or break your business. Learn more at vanta.com/lenny. And as a listener of this podcast, you get $1,000 off Vanta. That's vanta.com/lenny.
One thing I mentioned recently on the podcast, I heard, speaking of the code that you have of humans writing code, data labeling companies are buying code that was written before 2021, 2022, before AI became a thing, as very valuable data.
Original human code. Yeah, exactly. That's exactly right. And it's so interesting that that's exactly the kind of code used to build this model.
Well, what's interesting... So, I want to clarify there. I did not have a human write the code all by hand. Because I actually think that that's sort of... it feels silly to me. I don't really care, because I know if an engineer is not using AI, I'm not going to work with them. I don't really care. It's sort of like, am I going to race a human against a car? I probably wouldn't do that. But I would race a human in a car versus another human in a car and say which one's better. And in this case, the way the benchmark is structured is, yeah, these human engineers used AI, but they
used it in a way that I could not, cuz I didn't understand it and I didn't have time and I didn't really want to go in and try to understand the code base, to be honest.
And I think that's a really important thing when we think about benchmarks: AI is a broadly distributed technology that any human can use, and when we are benchmarking against humans, AI against humans, we're actually really always talking about one human using AI versus another human using AI, cuz AI doesn't use itself. It may be able to in this slightly somewhat recursive way, but in any real use case, there's always a human pretty close to it making sure that it's working.
Okay, I'm going to try to wrap up our first bucket. There's so much to talk about. I've made a little list of things that I think people should do based on your predictions to be successful. We'll talk about this at the end, too, but just a few things. One is start using Codex or Claude Code more and more for the work you're doing, and especially the browser use tools inside of it.
Two is allow agents to use your products. If you're running a SaaS tool, make it easy for agents to be a user, essentially.
Three is start thinking about some Slack bot that you can work with, like try out tools. I know Slack has their own Slack bot that I think is really good, too, and I haven't played with it, but people really like it. So, look for, I guess, a tool that could become the AI agent within your company.
Buy SaaS stock ASAP. Not investment advice.
I think that's totally right. My slight tweak is when you're thinking about building your software for agents, the current model is I'm building a CLI that an agent uses, but they're using it in a sort of like, I delegated a task to the agent and the agent's using the CLI. And where I think it's going is you and the agent are using the app together. The agent's probably using the CLI, but you're using the web interface, and they both need to be in sync. And that is, I think, a new challenge that's really interesting.
Awesome. Anything else before we get to our next category? Buy SaaS. That's the title.
Oh, man. Okay. So, the second category of predictions is around just the shape of the work that we're going to be doing is going to change. What do you predict? There's all this interesting stuff in terms of the shape of work. Once you're in this land where you've got, you know, async agents off that you delegate work to, then you've got your Codex, Claude Code work surface, that starts to happen. So,
one thing that we see a lot internally, and you also see this in the big model companies, is the number of pull requests that you get skyrockets. You know, we have people in consulting or in ops roles or whatever, or editors, just making pull requests.
And hey, that's really cool, and it's a very different shape of work where you can expect that a higher percentage of your company or your users are going to be doing things that previously only technical users could do. And what that does is it creates all this pressure on the other end for the people who have to deal with all of the new code, for how to deal with that.
And so, I think there's a lot of interesting things that happen with that. So, for example, OpenClaw, I mentioned that earlier. Pete gets like thousands of pull requests a day on OpenClaw, and then he just spins up like 50,000 Codex instances and then sorts through them and then merges like a thousand of them. It's really crazy. I actually think that that's going to be more and more common.
It brings up a lot of really interesting questions around which pull request should you merge? And you know, whenever you add capacity in one part of your process, it breaks things. It used to be really hard to build things, and now it's very easy. So, the point is not, can we build it? It's like, would it make sense with the rest of what we've built? And how do we keep a sense of a coherent whole? And also, what do we delete? I think Anthropic does this really well. They delete a lot of stuff from Claude Code to make sure it's not bloated.
So, I think there's a lot of that going to happen. On one side, non-technical people can do technical work, and then technical people are in charge of making sure that that work gets into a product or into a process in a cohesive, coherent way. And also, product people are going to be doing that, too. And I think that's quite cool.
Something I'm hearing from people is that now that everyone can do everything, like engineers can design, PMs can code, marketing people can ship stuff, there's just this confusion about what the hell is my job anymore. What am I responsible for, exactly? Am I supposed to be shipping stuff? Am I still a marketing person? And it's just creating a lot of confusion and uncertainty in the world.
I think that's for real, and one of the things that I think is special about Every is everyone is sort of a generalist and really loves having their fingers in a lot of different pots, or whatever the metaphor is. I think that'll probably settle down at some point, and it'll feel more normal. Marketing people are still going to do marketing, even if they're touching the website. That's just part of marketing now. But I also think that you can get a lot further being a generalist now, and that's really cool, especially for smaller companies.
The other thing that I think is interesting is there are definitely some new job roles that are a thing. And the thing that is becoming really clear is the whole forward deployed engineer concept, I think, is for real. And it comes out of every agent needs a human. Even if you go to the big model companies, they have these agents that run internally. They have teams of people that run these agents, you know? And I don't think those teams are going away. The models are going to get more powerful, the agents are going to get more powerful, and the number of agents is going to grow, but people are still going to manage them.
And so that looks like a very specific kind of person. And you know, we have a couple of those people internally here, and it's the people who are in charge of making sure your agents are working and doing the right thing. We also do consulting, so we lend that out to people, and I think that's a big thing that people want, and it's another one of those places where you're like, "Hmm, automation was supposed to take away jobs, but it looks like it just created one or many."
And there's a specific type of engineer that really loves this. You know, Nitesh, who fits this, he's an AI engineer and he fits the sort of forward deployed category, and he's on our team. He spends most of his time actually talking to one of our agents in Slack. We have an agent internally called Claudie, which runs our whole consulting practice. And he spends a lot of time in Slack. There is code, and he is using Claude Code and other things like that, but a lot of it is just talking to it and being like, "Why did you do this dumb thing? Let's fix that, you know?"
And so there's certain kinds of engineers that I think love that and love having their hands on the latest thing, and also love being in a workspace, and it looks a bit different than building more traditional software.
And your sense there is we're not near a place where these agents don't need a human. You said that so many times now, that agents need a human, and there's kind of the setup part and then there's the maintaining it forever part. It feels like both are important. What I'm hearing is this is going to be a job for a long time. AI is not going to get smart enough to fully automate it for a while.
Yes, I'm simultaneously extremely AI-pilled and very bullish on humans and the role of humans in making sure that AI is working well.
Interesting. Okay, so the two kind of buckets here that you're talking about: one is, the way I hear what you described earlier, the pace of shipping software and everything is just increasing, which also means there's so much more work reviewing all this sloppy output. I was just talking to a data science friend and he was saying how his data science team, their job used to be do analysis, answer questions, see if this experiment was positive. Now everyone's doing that and they're sharing the results, and they're like, no, this is not correct, and most of their job is now reviewing bad data science work.
Which is a problem, and the same thing is happening with engineers, and it means that you actually need data engineers for this and you need data scientists, and it means that you haven't set up the appropriate systems or agents to help you with this. So the way that it works inside of the big model companies, for example, at least one of them has literally a data science bot that every single person in the org can query, that is hooked up to their data warehouse, that knows who's who so that it knows at the warehouse level who has permission to access what.
And so all of the basic questions, because there's a team that sets up this bot, all of the basic questions that people might want to ask that it might get wrong, they're constantly making sure it's getting it right. And so the data science team doesn't have to answer all of the [ __ ] questions, because there's another team building an agent that is set up to do that really well. But if the team didn't exist, the data scientists would hate their lives.
Yeah. It does though make the job maybe less fun, cuz you're just sitting there, you know, gardening people's sloppy work versus
What I think is it can actually make the job better, because for the data scientist, you're now not dealing with all the silly requests. You're dealing with the deeper questions that are harder, with the team who's dealing with all the basic requests building an agent to do that. It's like filtering all that stuff out so you can focus.
Here's a question I've been thinking about. I was not planning to talk about this, but it's something that I've been thinking about. So the question is, which product tech role is the least changed now? So engineers, 100% of code is AI now. It's a completely different job. Product management, a lot of the PRDs you don't have to write as much. You can ship code. You don't have to wait for people. Design, the whole design process is dead according to recent guests, there's no time to do the whole design process, very different role. Data science, very different work now. There's marketing, there's sales. So here's the question. What do you think is the least fundamentally changed role so far?
Well, one interesting thing is, I don't know if this counts, but CEOs and investors, it seems still very, very optional whether or not they use this stuff.
Mhm. It seems that way. I think the opposite is actually true. My experience, we do a lot of this with senior executives and senior leadership teams. My experience is that your company's only going to go as far as your CEO goes in AI, and it's not something you can delegate. You have to have your hands in it, cuz otherwise you don't have an intuition for it. But for a long time it has seemed like, yeah, that's something that the people who are doing the work have to do, but I don't have to do that. I'll just tell them what to do.
And so I think if you're a CEO, you kind of can get away with your day looking very similar. I think that will change rapidly at some point, where it'll be like, oh no, I'm way behind, but for now, or maybe even middle managers, those kinds of people, I think it's fairly similar. I think maybe sales, because it's so, so personal.
That's, yeah, that's my vote. You know, it's sort of creeping up in the kind of BDR, like we can deal with a lot of BDR type queries. You're only talking to people who actually want it. And for sales, it's so useful to do research.
One of my favorite Codex experiences is we're hiring a head of L&D. And we always put out a job post, whatever, but I was like, I feel like there's this company called General Assembly in New York, and they've done really good technology education for a long time. And so I was like, I feel like someone who worked at General Assembly and is now into AI would be really good, and I just literally typed it into Codex and then went off and was doing something else, and I came back and it found the perfect guy. He worked at General Assembly, was an instructor, is super AI-pilled and follows me on Twitter. So I just DM'd him and then I had dinner with him. And it's like, that's crazy. You know, that would have taken so long before. And super valuable for sales, for recruiting, all that kind of stuff.
Yeah, sales is where my mind went. The top of funnel, AI is helping a lot with sourcing and qualifying, things like that. It feels like the work of a salesperson is not fundamentally different. And customer support has fundamentally changed. So that's interesting, sales. So far so good for those folks.
Yeah.
Okay. So maybe just summarizing some of the predictions in this bucket of just the shape of the work, how it's going to change. What I'm hearing so far is it's going to be a lot more reviewing of other people's output as a part of the work. And then two, there's going to be a lot of almost babysitting of AI agents to make them do the thing you want them to do, forward deploying and then just gardening them along the way, making sure they continue to do their work. Anything else before we get into our third bucket?
I would sort of split it into less babysitting agents and more your forward deployed team is trying to build a whole system that makes it so that people who have less knowledge can use that system without doing something dumb. And that's a really interesting engineering challenge. I think babysitting kind of makes it feel like you're just waiting for it to [ __ ] up and then fixing it or whatever. And that can be the case, but I think a lot of it is just an extremely interesting engineering challenge of building a system to enable everybody else in the organization to do what used to be a technical job.
And then if you're not one of those people, like you're the data scientist or whatever, you can go a lot deeper with AI into really important questions that eventually probably filter into the work that the forward deployed engineering team is doing, but is more generative and more new, and you're dealing with harder questions.
One last thing that I think is really interesting is I think that we will be reading way more AI-generated writing in documents and emails, and we will like it. And I think we are already doing this in coding, where we read plan documents. I don't want an engineer to handwrite a plan document. That would be very silly. It would be obviously silly.
And I think the same is true. You know, when we did our quarterly planning for Every at the end of 2025, we did it all with Notion agents. We had really one Notion agent, and then we had a top-level company strategy, and then we had everybody in the company just talk to an agent, and it asked them about what happened last year, how did
it go, what were your goals, what do you want to do this year, what are your metrics, it pushed back, and then it was like, how does this relate to the overall company idea? Like all that kind of stuff. And then I got these like incredibly good AI-generated like strategy reports or plans, like quarterly plans for each part of each team. And then I could go in and be like, okay, who needs to talk to each other? Like which teams need to talk to each other that like don't know they need to talk to each other? And, you know, which one of these is like actually low quality or which one of these is high quality? Like all that kind of stuff makes it a lot easier to process.
And I see that all the time now. Like I consistently get AI-generated stuff and there is a difference between an AI-generated document that's slop and not. And the slop one is it took them less time to make it than it takes me to read it. And they don't stand behind every line. So my expectation is, if you send me an AI-generated document, I think that's great. And if we talk about it and it's clear you have no idea what's in it, like big no-no. Not allowed to do that.
And I think this aversion to AI-generated stuff, that will go away because the kind of strategy document that GPT-5.5 can write when it's directed well by someone on my team is way better than like them just like dinking and dunking like their fingers on the keyboard. Right. Like most people are really bad at writing strategy documents. So, the bar is low. Yeah.
And same thing with email. Like most of my email is written by GPT-5.5 and Codex right now. And I honestly would prefer it to say that it's coming from GPT-5.5 and I may change it to do that. But I had this experience the other day where I had to send an email to one of our investors and I asked Codex, like, go do it, and Codex knows to ask me and it usually does, but this time it didn't. And it just sent the email. And I didn't look at it at all. And I was like, [ __ ]. And so, I went to my sent and looked at it and I was like, oh, this is exactly what I would have sent.
And so, it's pretty close to that a lot of the time. It can be like a little over formal and there's a couple things that it's just, when you really think about it, most of your email is kind of, it's not, it's kind of rote. It's kind of prosaic. It's kind of... I definitely want to be the one to think about what it should say, like what it should say, but the actual sentences don't matter that much to me usually. Sometimes they do a lot. And this is coming from a writer. Like I care a ton about writing. I think that human writing is incredibly important. And I expect we only publish human writing. Well, actually, we publish a mix of human and AI writing, but we always label it. Sometimes it's nice to have an AI co-author on certain things. I absolutely think that human writing is important and I think that the reaction or the aversion to AI writing is silly.
It's such an interesting lens on that because when people think about AI writing, I think about social media and videos. And your point is internally, if you're just like working on planning and documents and email and things like that, like that is much less scary that it's AI written. And to your point, people are already doing this. You almost prefer it a lot of times cuz people are really bad— Totally.
We have this too for external stuff. Like we publish all these guides and the guides are often agent-assisted and the agent is a co-author and they're intended to be read both by humans and by agents. And that's because like, if you're writing a huge informational thing, I mean, you do this all the time. In order to like really apply it, the best way to do that is just like have your agent ingest it and remember the next time I'm, you know, doing pricing to like remind me of this guide and we'll go through it together or whatever. It allows you to operationalize the ideas much better and it allows you to go much deeper because agents can read like 10,000 pages in like a second. And so you talk to the human about the story and the stuff that matters and the core ideas and the agent has all the details that it can then apply for you when you need it.
Awesome. Anything else in this category before we get into our final category? No. Okay, let's do it. So, the final bucket is just who will be successful in this AI future that we are approaching slash what should people be working on to be successful in this next year or two?
I am super super bullish on PMs. And I know that your audience will probably love that. But my anecdotal case that has convinced me of this is we have this guy internally, his name is Marcus, and he runs Spiral, which is our writing app. Marcus is a PM by training. He previously ran Axios's writing product and was a PM and had a big team and it got to, you know, tens of millions of revenue and ARR. And he took a year off that job and just got super AI-pilled. And just learned how to use Cursor basically really well. Now, I think he uses Claude Code, but he was extremely Cursor-pilled for a long time.
And he's, I would call him like lightly technical. Like knows what a database migration is. Like if he has to look at the code, I think he can understand it, but we never could have hired him to do this job even a year ago. But the coding models have gotten good enough that he can pair the kind of technical knowledge he does have with his really spiky product sense and sense for writing and sense for users. And it's so dangerous. Like he ships faster than almost anyone on the team and he has such an eye for every single user, every single conversation, like what does it mean and how do we collect it into a story about like where we want to go next and what are the issues we need to fix and like all that kind of stuff. And I think that he feels liberated cuz he doesn't have to organize a whole team of people to do that. He can just do it. And it's super impressive and it makes me very, very bullish on any PM who gets like really AI-pilled. Music to my ears, Dan.
You're making a lot of very happy listeners here. I've been saying this for a long time, too. It's just like the skills you need to build are the things, like the building out is done for you. What do you need to be good at? Figuring out what to build, figuring out if it's great, figuring out what problems to solve. So, I love that you're actually seeing this come to fruition. I really believe it. This could be the highest rated podcast episode of my whole podcast. There is going to be like—
Hell yeah. It's going to be okay. SaaS is back, PMs are back, you know. This is the most contrarian episode I've ever done. Oh my god.
So, okay, so the other people that I think are going to be like super super power people, and again this is cuz we see this internally, is full-stack designers. If you're a designer and you're in these tools all the time, you're so used to, okay, I make this beautiful interaction and the engineer like just doesn't want to do it or it doesn't like happen the way I think it should happen or, you know, there's all this stuff. And I see so many designers for us internally or externally where they now feel so empowered to like go build stuff cuz they're like, I have all these ideas to make things look amazing and these interesting interactions, and that's the exact thing that it's really hard to do with vibe coding because it just all looks the same so it all looks like slop, and they can make stuff that looks so different and now they can actually build it.
And what you see when we work with them internally is now they're just making pull requests. Like they don't need to hand it off as much. Sometimes they do but like a lot of times they just make pull requests and it's like the thing is built and that's it. And I think that's incredible for the way that companies work but there's also a huge opportunity for those people to become much better and like start their own thing cuz they can make stuff now. And I think designers are such creative people and I think AI is like a super tool for anyone like that.
I so agree. Even though there is Claude Design, there's all these AI designing tools, like once you see it, you're like, that's definitely Claude Design. And the creativity, to your point, it just feels like it's going to be more and more valuable to stand out from all the slop that people are shipping and launching constantly. So, I completely agree. It's interesting that designer roles, I do research on the job market, and interestingly designer roles have not grown in a while. So, I'm waiting to see if that becomes a big trend, just like we need more designers. Hm, that is really interesting. We'll see.
Yeah. We'll see. We'll see. That might be a way to predict this, is are people hiring more designers? I don't know. That is interesting. Yeah. All right. So, that's, so PM, designer thriving. PM, designer thriving.
I also just think generally the AI job apocalypse is not really a thing. Absolutely, we see companies starting to reorganize and I think that makes a lot of sense. I think to be honest a lot of the reorganization you can say it's AI, but it's like we over hired and like the company's not doing as well and all that kind of, it was like coming and this is a good excuse. But the like mass unemployment thing that I think like some AI CEOs are talking about, like I think that's not going to happen.
The pattern that I see so far, and again, I don't have a total crystal ball, but I do feel like we've seen enough of the new model drops to like have some sense of how this is going, is that what a new model drop does or what models do in general is they make yesterday's human competence cheap. So, what I mean by that is they ingest all this data of what has happened already and they make it really cheap to deploy that in whatever situation you want as your own, right? And what happens then is this is a new power that everyone has. So, it gets adopted super rapidly and suddenly that stuff is everywhere. It's like suddenly anyone can make a landing page, there's new landing pages everywhere. Suddenly everyone can write, there's like slop tweets everywhere.
But what's interesting is because it's all coming from these models and everyone's using basically the same models, it all looks the same if you use it in the most default basic way. And so, it becomes commoditized. Like it's not valuable anymore. And what humans do is we sort of go in there and we're like, "Yeah, we have all this like frozen human competence from yesterday. How do I use this to like make something new and interesting?"
And I really think that structurally, because of the way the models work, because of the financial incentives of model companies to like make them compliant and aligned, structurally, they're always going to be trailing behind those people who are taking the models and using them to make new expertise or make new things that haven't been done that way before for their very very particular situation. And that stuff is going to get incorporated into the models, but again, it will create room for people to push further ahead. And I think that you see this in a small way in like pretty much all the jobs, is like engineers. Suddenly, everyone's an engineer. That doesn't mean we fire the engineers. There's like way more demand for engineers cuz you need the engineers to like figure out, "Okay, this is all slop. How should this actually go in our code base?" And I think that's something that the benchmarks rising doesn't really capture. And it feels like a thing that will take a long time to change.
People may be hearing in this prediction here of just, "Okay, the job apocalypse is not going to... People are not going to be all fired. There's going to be human jobs remaining for quite a while." It may be almost too comforting because you probably have to change the way you operate to still have a job in the future. Do you have any sense of just like, "Here's what you need to do to not be one of these layoffs"? Yes. And I think that is actually super important.
The only thing you need to do is ride the models. And that means use them for whatever it is that you do. You know, we've talked about how Codex and Cowork are becoming a sort of standard operating system for work. If you're just doing that and when new models come out, you're trying them and figuring out, okay, how can I, now they're new powers, how can I use them, instead of just being like, I'm going to like try to ignore it cuz it like makes me afraid, which I think is honestly, it's rational, it's a reasonable response. And also if you ride on top of them, they extend your powers in a way that doesn't leave you behind. Like you're part of the future and part of the way work happens, and I think that we're going to need people doing that for a very, very long time.
I like this term ride the models. So, what, like, say a new model comes out, what do you think someone, say working at, I don't know, Salesforce. Say a PM at Salesforce, what should they do to ride the model?
Well, one of the things that's really interesting is a lot of companies like handicap their employees from even doing this because like, I don't know if you can use the latest models at Salesforce, you know, like a lot of times you have to wait or it's, you know, whatever. So, maybe you have to do it in your off time. But the thing that I really like to do with new models is play. And there are certain things where I know it can't quite do it yet, but when a new model comes out, I like always turn the rock over again to be like, can it do it now? You know, so, it could not do the senior engineer benchmark last time and I turned the rock over again and now it's at 60 out of 100, which is like really good.
So, the way to ride the models is like not one specific thing cuz they're always changing, but it is to be curious and playful, to apply the new model to whatever it is that you care about, whether that's your job or something outside of your job, and to keep turning over rocks because it may not work now, but it may work eventually, it probably will work eventually, and the way that you use it matters.
So, what's really cool is that I think people think of the edge of AI as being in San Francisco. And I actually don't think that that's where it is. I think the edge of AI is wherever AI meets like a real human doing something. Because the people in San Francisco, they're making it, but they don't actually know a lot about how to use it. They don't know, or at least they don't know everything about how to use it. They need to see how other people use it. And so whenever a new model comes out, you get to be one of the first people in the world to discover what it might be useful for. And like that's a new discovery. And I think that's why, for example, we're in Brooklyn. But I really think of us, and I think we are like quite far ahead of people in San Francisco because we just use them for everything. And if people do that consistently, I think it's going to be very hard to lose.
That is one of the most amazing things about AI right now is no matter how much money you have or little money you have, you have access to the most advanced AI model. Like it's not free, so you need some money. But like, and you can get it immediately when it comes out. Maybe the only people that have an advantage are the people working at OpenAI or Anthropic. But otherwise, it's just like available. I know I was at their event with you, their code event with you last week, or a couple
weeks ago and they're all using Mythos and I'm like, "God damn it." So annoying. [laughter] But I think that's totally true. If IBM had invented AI, you can bet it would not be like this. And it would be a bajillion dollars and only the top companies could use it and they would be using it in the weirdest, most uninteresting ways.
And it's really important that AI was built in America and in the Silicon Valley culture that's like we want to make intelligence too cheap to meter. That's not the default stance. And it means that everyone has this broadly accessible tool that they can use and I think that's amazing.
It's such a good point and interestingly it's also created the fastest growing companies in history, the biggest companies in history.
That's true.
Not the way to
Those Silicon Valley guys, they're smart.
If I zoom out on the conversation, it's really interesting. There's these two sides to the coin. One is so much is not changing. SaaS continues, jobs not disappearing. We're still emailing each other. We're still working in Slack. A lot of the work not changed. On the other hand, every role transformed. Engineers don't write code. PMs don't write PRDs. Design and design, you know, it's so interesting how much has changed, how much has not changed. I don't know. It's interesting that people think it's going to be this whole new world, but in many ways it's okay. It'll continue the way it is with a lot of stuff around the edges.
That's how I feel. I'm simultaneously so excited and it feels like everything has changed. And I'm so bullish on it and the progress that we're going to make and all that kind of stuff. And yeah, I just feel like there are these things where they're going to be pretty similar to how they are and that's probably good.
And I think generally our intuitions about the future, the model that I have of what our intuitions are about the future is the intuitions that people had in the Middle Ages about what happened at the end of the horizon, you know, it's like are there dragons? Does it drop off into nothingness or whatever? A lot of people have a lot of deep intuition that there's something terrible going to happen over the horizon. And also some people are like there's something incredible. It's going to change everything. We're all going to be happy as a utopia. And what happens is you get there and you're like, there's some really cool things, there's some not cool things, and it's just another horizon. And I think that's the way to think about the future. And until you get to that place where you're starting to see it, and I think we get to see it cuz we get to see it internally all the time, it's important not to let your mind get away from you and being like, this is going to happen and this is going to happen and whatever cuz you're going to tell a story that sounds so real in the moment, but later on you're like, actually it's much more complex than that and somewhere it's sort of a both: everything has changed and nothing has. And once you get there, I think you're sort of starting to see like, oh yeah, this is a real thing.
Part of it is that the AI companies are very good at scaring us about what might happen in the future. And I think that's actually shifting. I think that they've realized maybe we should not freak everybody out about the dangers.
That PR strategy just does not make any sense to me. I do think that it's genuine, but it's so ineffective and I think it's also wrong.
Mhm. How about we end with maybe just a few things listeners should do to be successful over the next year with the way the world is moving.
Ride the models.
I would try all of your workflows in Codex or Cowork and see how that works. And if your company doesn't let you do it, on your own time, I would try out some of these agent products like OpenClaw or Hermes or for less technical people there's Victor, we have 1 + 1's. I would get comfortable with both of those ways of working.
And try to have fun. I think there's too much of I'm doing this because I have FOMO, like I might lose my job or I might miss out on this big thing or whatever, and the best way to actually figure out interesting useful things to do with AI is to do something enjoyable.
We had Nikhil Singhal on the podcast and the way he described it is you got to find your moment of joy with AI. Once you find like, "Wow, I can't believe AI did this for me. This is awesome. I'm going to keep building stuff."
Yeah, I agree.
If you haven't seen that yet, then it's just like try solving it. The thing I hear a lot is just find a problem in your life or work and see if AI can do it. Go to Lovable, go to Claude Code, go to Replit. Try to build the thing and often it's like, "Holy [ __ ] this is so cool."
Dan, is there anything else that we haven't covered? We've gone deep on so much. Is there anything else you want to share? Anything else you want to predict or just say before we get to our very exciting lightning round?
I think we covered it. We did a lot. This is awesome and I'm very excited to see how well or poorly I do in a year and I hope that you hold me to it.
We're going to have AI score us. How about that? We'll look at the world like a dance prediction series.
Well, with that, Dan Shipper, we've reached our very exciting lightning round. I've got five questions for you. Are you ready?
I'm ready.
What are two or three books that you find yourself recommending most to other people?
Obviously Annie Dillard. Everyone at Every has to read The Writing Life. When you join you get a copy and you have to read it. You only have to read the last chapter though. I think the last chapter is incredible and it is at the intersection of writing, technology, and the future and its relationship to the future and to time and I think that's everything about Every wrapped up into a very tight chapter. It's so good and I think Annie Dillard just generally is fantastic.
What else do I recommend? I'll just tell you a couple things that I've read that I really liked recently, and whenever I like something I always just tell everyone about it. So, I have recommended these a lot. I've been reading... One of the things I learned, which I didn't know, is Churchill's a really good writer. And he has a whole history of World War II that he wrote, and it's a combination history and memoir. And I think that's so cool because he was there, you know, he did it. And there's something about what we do at Every, I feel sort of kinship with that of like we're building stuff, we're writing stuff, and it's very rare to find people that also do that. And so, Churchill's history of World War II is fantastic. I just finished the first volume. I'm on the second volume. The Nazis just invaded France. It's very captivating stuff.
So, that's one. I've also been on a little bit of a quantum physics kick recently. AI is actually very good for quantum physics if you get into it. And there's this book called The Rigor of Angels that I just finished, which is a history of ideas that relates Heisenberg, who has his uncertainty principle, Borges, who's an Argentinian fiction writer who wrote a bunch of great short stories that are actually starting to get a lot of play now cuz they're very AI-related, and Kant. Very cool, super mind-blowing. Lots of interesting overlaps with AI stuff. And yeah, highly recommend.
I feel like we could have a whole podcast episode about your reading and books you recommend. I know this is a passion of yours. My current obsession is The Power Broker. I think we talked about it when I was visiting you. It just never ends, but it's surprisingly compelling to read through the history of New York.
Okay, second question. What is a recent movie or show you've really enjoyed, if you have time for TV?
So, I've been watching a lot of basketball, so that's one. I became a Knicks fan this year, so that's really fun. But I recently watched this, I guess it's a mini-series documentary called The Dark Wizard about this guy Dean Potter who was Alex Honnold before Alex Honnold was Alex Honnold. And he just has this very extreme personality where he's free soloing everything and then he's, you know, base jumping in a wingsuit and stuff like that and it's sort of exploring his psychology and what happened to him. I don't know. I kind of like stuff like that. There's another one called 100 Foot Wave about big wave surfers. There's something about that, I guess it just reminds me of founders or whatever, but The Dark Wizard, highly recommend.
Is there a product you recently discovered that you really love?
Codex. It's the best. It's really good. It's really good.
Do you have a favorite life motto that you often come back to in work or in life?
Yes, I have several. The core one that I wrote for myself in college was do things worth writing about and write things worth reading.
And then there's this guy Rob Brezsny who's very popular in, you know, the AI meditation overlap discourse, which is also a big thing. And I really like him. He's dead, but I think he's amazing. And I listened to so many of his talks and there's this one talk that he gives where it's just one sentence, but he talks about when you're dealing with stuff that's hard, what you want to do is be able to relate to it from a position of spaciousness and strength.
And there is something I think really interesting and important in that. A lot of the meditation discourse or just generally how do you deal with hard things, it's a little bit more like the David Goggins, you just got to go for it kind of. And sometimes that can work. And also I think sometimes when you're dealing with things, so for example when you're dealing with I'm super afraid of how AI is going to, you know, change my job, it has been very helpful for me to be like, am I coming at this from a vantage point of spaciousness and strength, and if not, can I get there? Because it will be much more productive for me to deal with it from that place. And that has been very, very helpful for me.
Wow. I love that. Well, our final question, just on the theme of this conversation. Curious if there's an AI tool that you think is still kind of underrated that you're just like recently... I mean, people are sleeping on.
Codex. I hate to say this but I have to because anyone who knows me... We were at this conference recently, an Anthropic conference, and I'm telling Boris and Kat from Claude Code, you have to try Codex. It's just really good and the things that you can do with it are so different, especially if you're using it with the Anthropic browser to do things like your emails or check analytics or anything like that. It has completely transformed the way I work and I would be doing you a disservice if I was searching for something else because it is that good.
Damn, that's wild. Do you feel like Anthropic can catch up or is this just like well
No, yes, I think they can. Like I said, I think it's going to be a horse race and different people will be ahead at different times, but I think right now OpenAI has gotten back the mandate of heaven a little bit. It was a rough couple months, like 6 months or so, but I think they're back.
Interesting. And you'd switch if one became more
I would. I would. It's funny. People are like, "Oh, are you sponsored by OpenAI?" And I'm like, "No, I just talk about what I like." I was super loud about Claude Code when that was the thing I really liked. And I'll just say what I like when it happens, you know?
And to your point, there's a lot of value in using both for different things. So people
There is. I switch back and forth. I truly do still use Claude a lot.
Yeah. Such a big market. Well, Dan, we did it. We went through so much. I can't wait to revisit this in a year, slash get this out so people can start planning for this next year. Two final questions. Where can folks find you and Every, what should people know, and then how can listeners be useful to you?
You can find me on X at Dan Shipper, S-H-I-P-P-E-R, and you can subscribe to Every. Please subscribe to Every, every.to, every.to/subscribe.
How can listeners be useful?
You know, have fun with AI. Seriously, it's super fun. It's not necessarily useful to me, but I think it makes everything better when people put their hands in it and just start figuring it out together rather than arguing about it. And so the most useful thing you can do is find ways to use it well in your life and share it.
Dan, thank you so much for being here.
Thank you.
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 · Updated