Kevin Weil

Today we’re launching our Physical AI Lab, led by UCLA professor and renowned researcher Bolei Zhou, who’s joining Coco as Chief AI Scientist. Bolei has spent his career advancing physical AI with 60,000+ citations and pioneering work. Now, he’s bringing that expertise to Coco. What makes this lab special is Coco’s fleet itself. Our robots run urban deliveries across the U.S. and Europe, navigating unpredictable environments (construction, pedestrians, weather, etc.). These are the kinds of real world challenges you can’t capture in simulation and that data is becoming the raw material for advancing physical AI. As we scale to 10,000+ robots in 2026, this fusion of research and deployment will define the next generation of physical AI and reimagine how goods move through cities everywhere. TechCrunch has the scoop - link in comments.

213

12 Comments

For 20 years, quantum was hype without hardware. PsiQuantum just changed that. Thrilled to see Pete Shadbolt and the PsiQuantum team raise $1B Series E at a $7B valuation, making it the most well-funded quantum company in the world. The round was led by BlackRock, Temasek, and Baillie Gifford, with participation from Macquarie, QIA, Morgan Stanley’s Counterpoint, SentinelOne’s S Ventures, Type One Ventures — and NVentures (Nvidia), who will also partner with PsiQuantum on GPU–QPU integration, algorithms, and silicon photonics. It was great to welcome Pete at RAISE Summit in July at the Louvre. His vision stood out: • Betting on photonic qubits instead of exotic physics. • Building a cryogenic semiconductor fab moat that makes replication nearly impossible. • Tackling the million-qubit fault-tolerance challenge head-on, by industrializing quantum chips at scale. Why this matters: • Compute is the new oil. AI demand is already straining global GPU supply. • Quantum is the natural next step in compute — solving problems beyond the reach of even AI-accelerated superclusters. • Applications in climate, new materials, drug discovery, finance could unlock trillions in value creation. Grateful to Tony Kim for the introductions and insights into the quantum space. At RAISE Summit 2026 in Paris, we’ll dedicate a major track to Quantum × AI — where frontier compute converges. The next decade of breakthroughs will be written at this intersection

227

7 Comments

GPUs have gotten faster, and models have gotten larger. But storage is still stuck in the past. Modern AI workloads need instant access to huge datasets. Instead, teams waste time on slow retrieval, idle GPUs, and costly pipelines. Archil is changing this. It delivers object-storage scalability at block-storage speeds. The platform allows for seamless, high-performance access to massive datasets. By eliminating cold starts and accelerating throughput 30X, Archil enables teams to fully utilize their infrastructure and move faster on training, analytics, and AI deployment. Archil’s founder Hunter Leath brings rare technical depth and operating experience from building Amazon EFS and optimizing Netflix’s cloud performance. This is a team building the foundational layer that AI workloads desperately need. I’m thrilled to lead Archil’s seed at Felicis with Nancy Wang and support their mission to reinvent cloud storage for the AI era. https://lnkd.in/gWueXhzz

65

2 Comments

While we haven't officially launched, can't help but be excited to see this great artcle in the @WSJ featuring LGND AI, Inc.(http://lgnd.io) and the opportunity for AI applied to earth observation data for all sorts of important uses. This is a huge and important data source that is vastly under represented in the current AI platform shift - and we aim to change that. Big stuff ahead! https://lnkd.in/ghd76mBB

89

9 Comments

Had a great discussion with Crunchbase on the "new normal" in the AI-era. As AI continues to expand the software pie, we're seeing larger addressable markets, more mega rounds, and higher mortality compensated by larger foundational companies. Check out the insights here: https://lnkd.in/g9NugRu3 #Software #TechTrends

67

2 Comments

Thrilled to back Dima Shvets and Alexey Moiseenkov on the launch of Mirai On Device AI. If you’re building AI-native products, this one’s for you. Mirai is an on-device AI SDK for iOS and macOS that lets you run high-performance models directly in your app with zero latency, full data privacy, and no inference costs. So instead of sending tokens to the cloud (and paying for it), you keep things local. Even better, you don’t need a dedicated ML team or weeks/days of infra setup. One developer can integrate inference, routing, and optimization in minutes. Key features: — Supports Llama 3.x, Qwen 3, Gemma 3, DeepSeek-R1 — Bring-your-own-model support — Tools for fine-tuning, evals, structured outputs & fallback — Integration in a health tech product with 100M+ MAUs Built by the team behind Reface & Prisma. Backed by scout funds of Sequoia Capital & Index Ventures, and angels from OpenAI, Hugging Face, ElevenLabs, Unity and us, at NP-Hard Ventures 🤩 On-device AI will be a major unlock over the next few years, and help us to shift the way people interact with technology. And Mirai is early, ambitious, and razor-focused on enabling this shift. Go try it out! Link to SDK and GitHub documentations in comments. NP-Hard Ventures // Paul Veugen // Micha Hernandez van Leuffen

72

13 Comments

⚠️ Founders - consider the below reasons to not use Forward Deployed Engineers (FDEs) to acquire your *first* customers. This signals (1) lack of PMF; (2) miscalculation b/w development and sales timeframes; and (3) mischaracterization of Palantir Technologies timing when they deployed their FDE model. 💡 At the early-stage, it's crucial to demonstrate ability to repeatedly acquire customers with limited resources - capital, team, & time - that supports the narrative of 'more funding will amplify my traction.' Bespoke solutions for one company demonstrates zero repeatability. These custom solutions will not carry over to other customers within the same cohort. 💡 AI accelerates engineering but sales, system integration, and stakeholder buy-ins are perpetually slow. This imbalance results in concentrating all your engineering resources into one company, ignoring other opportunities, and still not going live on an accelerated timeline. 💡 Palantir didn't start using FDEs until year 3 when they got their first *gov't* customer, not their first customer. Gotham was already built and had strong PMF. Gotham allowed Palantir to take GTM risk by using FDEs b/c if it didn't work, Gotham was humming in the background. TLDR: At the early-stage, repeatable customer acquisition is the holy grail. Don't let AI deceive you that it can accelerate human behavior like it has with development cycles. It C.A.N.N.O.T.

32

1 Comment

Grok 4 was just released today and it's already crushing SRE tasks. It’s not just us, the broader community has already noticed Grok 4’s standout performance on ARC-AGI, the gold-standard benchmark for fluid intelligence created by François Chollet. At Rootly AI Labs, we built a benchmark focused specifically on real-world SRE workflows. Grok 4 came out on top, beating even the most expensive models. We’ve started experimenting with it in production with our AI SRE assistant. It’s early, but the results so far? Seriously impressive.

77

7 Comments

What does 'commoditization' of LLMs mean for founders and builders? Why is YKK so successful if zippers are a commodity? Will Sam open up OpenAI model weights? What did Gaurav Sukhatme have to say about history books? They actually all answer the same basic question - what's next? I think we see 3 kinds of companies emerge: 1. core private model + product wrapper - OpenAI 2. core public model (free, strategic gift) - Meta 3. commodity model + product wrapper - Perplexity Dragonfruit AI is squarely focused on #3. The reason we see Perplexity and Glean get big so quickly is because they're building on the second easiest data type - text. But that's also where incumbents and new entrants can quickly catch up, before these companies lock up their markets. Over time, I suspect the moat will shift to companies that can deal with data types with other intrinsic complexities that AI can't simplify - images, then video, then real-time video. That's what Dragonfruit is focused on. (btw, the easiest data type is numbers; wall street already did quick work on that decades ago)

25

4 Comments

The NYT nails the current state of AI in this essay. It's exactly the insight we had 18 months ago and why we started HumanX. Indeed, I even talked about the bubble and the trough of disillusionment during this year's opening remarks. The hype train does very little to build the lasting value that I'm convinced AI will bring - we need to focus on the practical applications AI is bringing - TODAY. Eric Schmidt was mentioned in the piece and I will leave it here - he expresses the sentiment perfectly: "This week the longtime A.I. booster Eric Schmidt, too, shifted gears to argue that Silicon Valley needed to stop obsessing over A.G.I. and focus instead on practical applications of the A.I. tools in hand." This is exactly correct and precisely why HumanX exists - to help enterprises navigate their way through the AI forest to find their highest, best use cases for what is unquestionably one of the most important technology shifts of our lifetime. We just can't get distracted by the shiny baubles - awesome as many are - and let them detract from the incredible work and productivity gains enlightened enterprises are already experiences when used correctly! https://lnkd.in/eiH79qx9

62

5 Comments

Startup radar time! Meet another group of early stage Seattle-area startups: CaseGuild: Led by Vikas Rajvanshy, CaseGuild leverages AI to help legal teams wade through massive troves of evidence. GenSX: Led by Evan Boyle, GenSX is building a programming tool for building AI agents and workflows. GradPilot: Led by Nirmal Thacker, GradPilot gives students an "AI counselor" to help craft college essays. HA Arts: Led by Harshitha A., HA Arts blends generative AI and human creativity to produce commissioned artwork. Harte.io: Led by Christian Hammer, Harte aims to consolidate disparate software systems within a company’s tech stack into a unified model. Read more: https://lnkd.in/g_2CyBVP

89

4 Comments