We are entering a new dark age of cybersecurity. Devin is here to help. Introducing the new Devin Security: powered by Claude Mythos and frontier cyber models, plus hundreds of millions of dollars of credits to help teams find, prioritize, and fix urgent vulnerabilities.

This is directionally consistent with our internal data @cognition. Many tasks within software engineering are now "intelligence saturated." Once any model can do the task well, cost and speed are all that matter.

Karim after winning corporate cards and now going after the AI labs

Should hardware engineers trust AI?

A year and a half ago we priced @Ramp's $13b round, and our call then was that they would be a multi-product winner. Today we are co-leading Ramp's round at more than 4x the price - and our call is that Ramp will be an AI winner. I remember a conversation with @karimatiyeh a number of years ago where he said something to the effect of "the funny thing about investors is that they only seem to extrapolate the traction of our existing products, whereas we spend the majority of our time building products that are nowhere in an investor's model, so either we are terrible capital allocators or we are going to outperform expectations by successfully building additional products that serve our customers' unmet needs." As an investor, it is easy to get distracted by the trailing indicators - and certainly there is much to be excited about with Ramp growing 70% faster today than it was a year and a half ago. Yet our excitement about Ramp has more to do with the forward-looking indicators. We believe Ramp is one of the most AI red-pilled companies in our portfolio. They are building for their customers' future needs and will continue to set the bar for product velocity and rapid innovation. They are customer-centric and relentless. We are honored to call them our partners. @eglyman @faxherbert

Excited to welcome you to Georgia! @ZachBDell @GovKemp

Even a year ago I don't think I would have believed that PI models can run on such a huge range of different robots.

Cognition has crossed $1B in annualized revenue run rate. This milestone belongs to our customers. Here's how a few of them are building with Devin.

Great entrepreneurs surprise you to the upside. We first invested in Crusoe with a thesis that they would become the market-leading developer of data centers for hyperscalers and AI Labs. We believe more than $1T of market cap will be created at the data center development layer over the next 5 years as we undergo the largest infrastructure buildout since the Industrial Revolution. Yet over the past six months, we have had to update our priors. @ChaseLochmiller and team have leveraged their advantaged position and unique capabilities to successfully build a vertically integrated neocloud that we believe could become the lowest cost global producer of tokens. We are excited to double down and invest again in their Series F.

We first met @mjlee_2014 more than 5 years ago as a fellow investor and friend. Whereas many of our smart investor friends talked about the generational opportunity to leverage AI to transform businesses, Michael bet his career and moved with a speed I had not before witnessed to launch Sequence. I have met few people in my career who can think deeply like an investor while executing quickly and strategically like an operator. Michael is one of those rare examples. It was an obvious decision to be a day 1 investor with Michael, but he has dramatically outperformed my high expectations of him. So naturally it was no surprise to hear that he would target the largest AI for transformation take private in history, and even less surprising that he would pull it off! Congratulations Michael and team Sequence on this incredible milestone. Can’t wait to see all you will accomplish with Baldwin. Looking forward to working with a great group of partners like @MichaelDell and @ZachBDell in the years and decades to come

We just wrote our biggest check in Avenir’s history.  It’s our sixth investment in @cognition, at 10x the price of our first.  After two years of watching @ScottWu46 build, our conviction has only grown.  Proud to co-lead the Series E.

Right now, router.com’s traffic is compounding at roughly 34% per weekday. It’s early, but the product is useful and the growth curve is as steep as any I’ve seen.

💪 @joshim5 and @jackdent on a heater bringing the largest and most innovative pharma companies like Bristol-Myers Squibb, Eli Lilly, Pfizer, Novartis and argenx into the AI era with @chaidiscovery

When we first invested in Chai’s Series A, we were drawn to the audacity and the ambition of @jackdent and @joshim5 who envisioned a future of AI one-shotting novel drug candidates. Fast forward just a year and a half later and that vision is becoming reality with some of the largest pharma companies in the world. While AI’s impact in enterprise will create winners and losers, in biology and healthcare, all of humanity will benefit from more cures and better treatments thanks to the hard work and vision of teams like @chaidiscovery. Excited to have friends like @mamoonha and @lachygroom on this journey with us. businesswire.com/news/home/2…

Vive la France! @karimatiyeh

Talk about talent density! Last night, Avenir hosted an AI dinner in Paris with some of our favorite founders and friends @ScottWu46 (Cognition), @andrewdfeldman (Cerebras) @oliveur (Datadog), Chris Clark (OpenRouter), @paraga (Parallel), @nikhilbenesch (TurboPuffer), @dylan522p (the self described shit poster and SemiAnalysis legend), @robertwachen (Etched), @goodwin_ml (Fractile), @annadgoldie (Ricursive), @agermanidis (Runway), @PhilipJohnston (Starcloud), @dan_lahav (Irregular), @G_Princen (Anthropic), @JacobWallenberg (Ramp) and Rohit Iragavarapu @graceisford , @DelahayeHenri @DavidPrilutsky. Great discussion on open vs closed source models, intelligence saturation, the inference explosion, the geopolitics of AI and the future of work

I remember distinctly my first visit to the Cognition house in Atherton a year and a half ago. @ScottWu46 and @russelljkaplan had amassed the most talent dense team of IOI medalists I had ever met, with some of the best engineers and mathematicians in the world. This cracked team was unequivocal in their belief that models would eventually be better at coding than humans, which seemed like a stretch at the time. They saw the future clearly and built the first autonomous and asynchronous coding agents - thereby creating a new market that we believe could become the largest in technology history. We ended the day with some of Scott’s legendary mental math games, and I drove back to the airport convinced that I had just met one of the most exceptional founders of our generation.

We’ve been thinking hard about @AnthropicAI and the white-collar productivity stack. Microsoft is the 800-pound gorilla, winning with a bundling playbook for decades. We made a deck with some primary-research-backed recs on how Anthropic can beat it. Link in replies, enjoy!

If your kid’s lemonade stand processes 0.5–1% of US GDP, then yes, that’s a fair analogy for @tryramp. Ramp’s data is useful for the same reason it gets cited at all: it is quite consistent with the revenue figures OpenAI and Anthropic release. If it weren’t, no one would care.