5 years we toiled, happily, in obscurity and assumed it would always be that way. That makes today's @NewYorker piece all the more thrilling. It will be framed on so many of our mothers' walls. Special thanks to @UnrealBluegrass and Dr. Paul Krawitz for sharing. @fortellresearch

Even with the recent press and stock run up, not enough people are talking about Bending Spoons' run. -Another acquisition this AM at 90% discount. -Using their recent IPO to generate broader brand awareness to become the landing place for SaaS. -Every VC now telling their struggling companies to sell to them -Simultaneously advertising how their talent bar is extremely high. Wasn't familiar with your game, Bending Spoons. Well done.

Been a very fun and eventful 6 weeks so far :) Devin is now powering engineering teams at the world's largest companies, including NVIDIA, GE , Citi, Mercedes-Benz and many more. Job's not done🚀🚀🚀

GPT-6 Astra is coming to Devin. On FrontierCode 1.1, Astra performs within 0.4 points of Fable 5 at a 64% lower cost. It also sets a new SOTA on our internal testing benchmark, generating more comprehensive tests, clearer reports, and better video evidence.

the Cursor/OAI drama feels like a pretty big opening for Cognition. still independent/model agnostic/has access to all the models turns out "we're not tied to any one lab" may be a pretty good enterprise pitch

Leaders Must Manufacture Discomfort @frank_slootman (Frank Slootman), former CEO of Data Domain, @ServiceNow , and @Snowflake, interviewed by @nakul (Nakul Mandan), @AudaciousHQ (Audacious) (Knuckle Up with Nakul) Summary: Frank Slootman is the only person to take three enterprise software companies public, and he thinks the CEO's job is to keep the organization permanently uncomfortable. Most leaders set goals they already know they can hit and keep people they have privately written off. Slootman's answer is to compress every timeframe, buy aptitude rather than experience, and attack doubt by acting on it. Take that seriously and you stop protecting a calm quarter and start finding out what the business can actually do. 1. Time Compression. The CEO job is confrontational because you spend it compressing timeframes. Slootman treats every meeting and hallway conversation as a chance to change the intensity, the urgency, the pace, and whether the team is aiming high enough. He opens exec conversations with "how do you think it's going," then keeps opening the aperture until the person gets to his conclusion on their own. Most of them work out a week later that they got their ass kicked. 2. Drivers And Passengers. Come home on a Friday night, look in the mirror, and ask whether it mattered that you were there and where you moved the dials. An engineer once asked Slootman at an all-hands how to tell which one he was, and the answer was "you better find out before I do." Passengers are usually articulate, well-liked, and never stand out in a negative way, which is what makes them hard to see. Every company carries some, and Slootman says that in big companies it becomes a disease that eventually gets treated with a layoff. 3. The Aptitude Purchase. Slootman's first question to an exec candidate is what they are innately, effortlessly good at, the thing everyone else marvels at. "I'm not buying your experience because I can get you experience. I cannot get you aptitude." The inverse question follows: what job is absolutely lost on you, what should we never ask you to do. Most candidates treat both as trick questions, and Slootman says many have never been asked either one. 4. Back-Channel References. An interview gives you a vibe, and a vibe is a feeling. The real signal comes from superiors, subordinates, and peers going back at least 5 years, and Slootman puts his time there rather than into a 15-person interview loop. His favorite reference question is which of those three groups the candidate had trouble with, since almost everyone had trouble with one. He likes hearing peers, because ambitious people who take ownership clash; subordinates is the answer he does not want to hear. 5. The Empty Seat. "When there's doubt, there's no doubt" was the tiebreaker on every hire, and Slootman used it on every decision, not just recruiting. Once someone is clearly wrong for a role, coaching is usually a fool's errand that costs you time and ends in separation anyway. Having nobody is better than having somebody who is not the right person. CEOs sit on mediocrity because moving makes them look wrong, and "people don't just watch what you do, they watch what you don't do." 6. The 125 Question. At Data Domain, coming off $45 million, the VP of Sales built a plan for $100 million that the board would have rubber-stamped. Slootman asked what he would do differently if the number were $125 million, listened to the answer, and said "well, why don't we just do that, then?" They hit $125 million by changing the assumptions rather than the resources. Goals are insanely powerful because people immediately start breaking them down, which is why a limp goal quietly wastes the business. 7. Heels And Tips. 99% of people do not lean in hard enough. Slootman puts it in ski terms: skis are built to be ridden forward, and people ride them on their heels. Push until evidence piles up that you are overdoing it, because with a good product the evidence almost always piles up that you are underdoing it. If the resources genuinely do not convert to yield, you have just learned something more important about the business. 8. Decision Velocity. Getting everybody aligned is nice, and getting the right people aligned is essential, so Slootman never treated consensus as a goal in its own right. The question is who is carrying the execution, because those are the people who have to buy in. Waiting is a bigger risk than going, and acting triggers energy and speeds up the learning that follows. On harassment and integrity violations he moves the same day, because the organization reads response time as the real policy. 9. Standards Are The Culture. Culture exists to serve the mission, and a high-growth operation is uncomfortable by construction. The useful question is which behaviors the mission needs, which usually means a tolerance for growing faster, risking more, and working harder than feels comfortable. Integrity and respectful interaction were non-negotiable at Slootman's companies, and violations got prosecuted, because consequences are how people learn a standard is real. Culture you are unwilling to prosecute stays a set of good intentions. 10. Manufactured Anxiety. Leaders should drive high anxiety through the ranks on purpose so nobody takes their position for granted. Slootman lives in anxiety even when the numbers are good, because believing tomorrow will be fine is how people go back to sleep. He reads the current AI moment as one giant wake-up call: you may have been comfortable before, but you probably should not have been. The question he puts to CEOs is whether they are actually processing it. 11. Evaporating Swim Lanes. Slootman changed course at Snowflake because the company's swim lanes were evaporating, turning a predictable, profitable selling motion into a mega market with unfamiliar boundaries. He compares it to Intel leaving memory chips, where you commit to the move before you can see the other side. That changes how you operate, because you now have to try many things knowing most will fail. He also handed the CEO job to Sridhar Ramaswamy while holding the largest individual stake, on the view that the company needed something he did not have. 12. Providence Follows Commitment. Slootman's answer to doubt is to attack it with everything available rather than wall it off, and he describes wanting to face his demons "for breakfast." In Data Domain's first year the product moved 25 megabytes a second and everyone told him it would never fly. They found use cases small enough to sell, did $3 million, and stayed alive long enough for Intel's multi-core gains to carry the product. "When you commit, providence commits as well," and when you hesitate the world does nothing.

How is @cognition working with Devin in Slack Code? 👀👇 Devin follows the work as it happens, picks up a task from the engineering channel, opens a PR with a demo video, and brings in a designer to finish the UI. All without leaving Slack or waiting to be tagged. The agent isn’t off doing its own thing. The whole team is right there, building with it. That's the beauty of multiplayer AI in Slack.

Proud to be a launch partner for Slack Code. When it comes to agents, etiquette isn’t a given. Devin replies fast and gets to the point, not bombarding your teammates. We are everywhere you work, and we are polite about it.