
anand iyer
@ai
Followers
44K
Following
47K
Media
2K
Statuses
17K
Managing partner Canonical · Venture Partner Lightspeed · Core Contributor Few Shot · Father · Husband
San Francisco, CA
Joined February 2008
Expect VC rounds to inflate further, as founders now raise just to stay in the game.
Per Wired, Meta is poaching researchers with packages up to $300M over 4 years (with 1st yr comp >$100m), and that they've made 10 such offers to people at OpenAI. For context, this is roughly what Meta's execs get paid per year:.• CFO: $23.5M.• COO: $25.5M.• CTO: $23.5M.•.
2
1
8
RT @TheCinesthetic: Since its release 11 years ago, just one hour and 31 minutes have passed on Miller's planet in Interstellar. https://t.….
0
20K
0
Apple considered building its own AWS competitor, spending resources evaluating the complexity and cost of cloud infrastructure. Meanwhile, @exolabs has already proved the power of distributed in-house hardware, turning idle Macs into an active, decentralized data center.
Report: Apple looked into building its own AWS competitor by @mvcmendes.
9
2
39
This is a @SuitsPeacock spinoff waiting to happen.
PSA: there’s a guy named Soham Parekh (in India) who works at 3-4 startups at the same time. He’s been preying on YC companies and more. Beware. I fired this guy in his first week and told him to stop lying / scamming people. He hasn’t stopped a year later. No more excuses.
2
1
6
RT @joshelman: I still remember when the tech world felt like a tiny little cottage industry. Publications like TechCrunch were just starti….
0
2
0
AI agents fail like "radioactive decay" with a half‑life that cuts success by 50% every set time of human‑equivalent work. And even if you survive that, the model still risks "context rot" - as context grows (especially past ~100k tokens), distractions and stale info seep in and
Is there a half-life for the success rates of AI agents?.I show that the success rates of AI agents on longer-duration tasks can be explained by an extremely simple mathematical model — a constant rate of failing during each minute a human would take to do the task. 🧵.1/.
11
2
8
Well said, @bgurley. Tokenization of private assets.
This is an excellent post from Tomasz. I’ll take it a step further: it doesn’t have to be IPO and tokenization. It will increasingly be just tokenization. Why?. • IPOs are expensive (~$20M).• The IPO bar keeps rising median software IPO now needs ~$200M ARR*.• M&A is.
10
1
5
Sovereign AI matters. Alex talks about how giant “super-assistants” can nudge our choices. Recent studies show that heavy ChatGPT use can blunt students’ higher-order thinking instead of sharpening it, and researchers have already measured how different chatbot feedback rewires.
The real AI battle isn't about who has the best model—it's about who controls your context. I just published a piece on why keeping these layers separate is the most important design decision of the AI era:
9
4
24
RT @NexusLabs: Testnet III has begun. Today, we're rolling out Testnet III across the United States, Vietnam, Nigeria, Russia, India, Indo….
0
895
0
RT @Ar_Douillard: NoLoCo: Gossip learning x DiLoCo = ❤️. NoLoCo ( from @gensynai proposes two contributions on top….
0
23
0