We're excited to announce that Cloudflare is building an experimental consensus service using a never-before-deployed consensus algorithm, QuePaxa. https://t.co/hJ7dZIP3u0
Super excited about this! I spent 6+ years on gVisor at Google, then recently joined Modal. It’s been an adventure pushing the frontier of gVisor’s capabilities and performance here.
@modal is one of the most advanced gVisor users, running 40M+ containers a day, so we care a lot about where the project goes. Glad to see it move to CNCF, and happy to be back as a maintainer.
@brandur yeah all that, plus tons of benefits gone, silent layoffs, hiring management from AWS et. al., folks not happy with their work at Stripe, etc.
Basically the slow burn into “bad big tech” that people used to praise Stripe for not succumbing to.
For fun, overly long (slightly bearish?) thoughts on Stripe —
Yeesh, I found out about a couple more personnel churns yesterday. Out of all the best people I worked with at Stripe, roughly half are at Anthropic and the other half are at OAI, with a misc few that fell through the cracks and went to startups. In absolute numbers, the extent to which this is true is actually funny, and a bit unnerving lol.
They're facing a perfect storm:
• Ant/OAI's engineering comp bands are so much higher than Stripe's that they more or less don't even overlap. i.e. The min comp at Ant/OAI is above (or about the same) as Stripe's max.
• Ant/OAI have both announced intent to IPO, providing a clear signal of future liquidity. The Collisons remain adamant in their intent not to IPO, and remind employees and potential hires of this on a regular basis (most recently, Aug 2026 investor letter). With every tender offer they've always been careful to insist that "no future tenders are guaranteed" and they made good on that this year, skipping what'd become a fairly reliable H2 tender.
• Stripe operates in the usual ultra-HCOL cities (SF, NYC), exactly where Ant/OAI also do all their hiring. But Ant/OAI are paying enough for the 52% tax and $7k/mo rent. Stripe isn't.
Stripe's processing payments for both Ant/OAI and as seen by their latest $159B valuation, gotten a major tailwind from that. But as we've seen recently from Ant/OAI who both literally reinvented Git (!!), it cannot underestimated the extent to which they'll rebuild in house. They've got the resources, the technology, and the grit to reinvent anything. I wouldn't be surprised if one or both companies have skunkworks microchip projects for god's sake.
Stripe's always prided itself on A-tier talent, but it's safe to say that their pipeline for that is closed. Is the plan to drop to B-tier/C-tier, or fall back on Europe or Asia? These are the types of conversations exec is having behind closed doors.
Talk about missed opportunity though! If they had IPOed during their fairly clear window (2025, early 2026), the market would have been extremely kind. These guys were the darlings of Silicon Valley and the stock would've been a rocket ship. Now I think I agree with the Collisons — it'd still go well, but they'd be upstaged by more impressive companies. Maybe private is better.
Stripe will be fine — it's still led by two of the best founders in SV, but it's a case study in how an A-tier company that can do no wrong and walks on water can slip to a B-tier company. You either die a hero, or you live long enough to see yourself become Paypal.
Imagine if everyone had the tools to shape intelligence itself. That day is today.
Today we launch the Trajectory platform - the easiest and most intuitive tool to post-train models of your own. The foundation for continual learning.
With just a few agent prompts, you get access to a full creative suite to train, eval, and deploy trillion parameter models.
On top of that, we believe in a future where you don’t just shape your model once, it also keeps learning on its own from the world around it.
We’re releasing two research previews.
1. Continual grading scales your taste, so your model learns what good looks like from far more than you could ever grade by hand.
2. GS1, a new algorithm that lets models learn directly from what they do in production.
We’ve now raised 60M, with a new round led by @Sequoia at a 300M valuation to democratize the most powerful technology of today - training models - so that the lawyers, doctors, and engineers of the world can own their own intelligence.
You know what good looks like, now go build it. Own your own value function with Trajectory.