My AI security research lab AISLE discovered 500+ zero-days with CVEs across 100+ codebases in a year => ~1% of 2026's open-source CVEs globally from a single team! On hardened codebases like curl, we've repeatedly matched or outperformed frontier AI labs in zero-day discovery

Introducing DeepRSI, a new industry standard for measuring recursive self-improvement. On preexisting benchmarks, tasks are contributor-based and short-horizon (~12 hrs). In DeepRSI, we push agents to work 48 hours on ML tasks actually derived from ongoing proprietary research.

Wrapping up my first month at @DiscoLoopAI (am on leave from @PrincetonCS and @PrincetonPLI ). Excited to work on using AI to accelerate scientific discovery and solve important problems. My first startup experience, and with an incredible set of colleagues!

We’re sharing new results showing that frontier AI models are able to outperform human experts on financial predictions for the first time. We built an environment for the challenging task of Earnings Predictions, finding that the most recent frontier models (Opus 5.5, Fable 5.1, Astra) with Samaya’s harness, meaningfully outperform expert consensus predictions on commonly reported earnings metrics such as company revenue, gross margin and more. We built this environment to have a “point-in-time” gate in the harness to ensure no information leakage, and created financial retrieval and data tools that would respect this gate and give the models access to the same information available to experts. We also adjusted for bias in expert consensus to create a hard, bias-corrected baseline, where outperformance has real signal. Only the most recent frontier AI models are able to beat this hard baseline, pointing to an important inflection point in AI capabilities in finance. Like with FrontierFinance, we see substantial headroom for further RL training in this environment — the models show large variations in error across different rollouts. We’re continuing research on using earnings prediction and other predictive tasks to post-train models to learn from their mistakes. Get in touch if you’d like to collaborate or try out an alpha version. Link to the full research below.

Proud to share that @phinity_ai is coming out of stealth today, announcing our seed fundraise and that we’ve reached 8-figures in annualized run rate. Chip design has been stuck in the 20th century for far too long. Now, for the first time in decades, a unique set of emerging factors is making it possible to fundamentally reinvent how chips are designed. At Phinity, we’re building toward fully autonomous chip design, with the goal of enabling custom silicon for thousands of emerging AI applications. Our team combines hardware experts, AI researchers, and infrastructure/EDA engineers to define a new way to design chips. We’re extremely grateful to @uncorkcap for leading our $5.2M seed, to @MoxxieVentures and angels including @JeffDean for joining the round, and to @pearvc for leading our pre-seed and backing us from the start. We’ve worked so, so hard to make this happen, and I’m proud to share our work publicly for the first time! You can read more about the announcement below. We’re hiring for ops and engineering roles. If you want to help build the future of chip design with us or are interested in learning more/partnering, please reach out to me or @sonyashijin .

Excited to finally share that @phinity_ai is now out of stealth! We're announcing our seed fundraise and that we’ve reached 8-figures in annualized run rate. We’re already working with some of the largest foundation model labs. Our mission is to build fully autonomous chip design so AI can help us search beyond the limits of human engineering and discover computational systems we would never find on our own. We’re fortunate to have the support of @uncorkcap, which led our $5.2M seed round, with participation from @Moxxie including @JeffDean, following our pre-seed led by @pearvc! It’s been an amazing few months, and I couldn’t be more grateful for the opportunity to build the future of chip design with such a talented team. You can read more about what we’re building and why in our full announcement. We’re hiring across all disciplines - if you’re interested in working with us, please reach out to me or @aadi_nash.

Very sorry to hear this news. 😥 I never had the chance to meet Margaret Hamilton, but her contributions to the flight software that got us to the moon are legendary. I love this image of her standing next to the printouts of the Apollo code.

We are excited to announce a new faculty position in AI for Biology at UC Berkeley, jointly between @Berkeley_EECS and @CCBBerkeley. Please repost & share this opportunity with outstanding early-career researchers in your networks. Apply by Nov 9, 2026:

The safety data for Waymo gets better and better. Comparison w/ human drivers on rate of crashes with serious injury: Latest data: 270M miles, 20X better Mar 2026: 170M miles, 13X better Before that (forget exactly when): 10X better

Luiz André Barroso’s work influenced me a lot in how I think about systems, and even about my career. I decided to write about his work with @JeffDean on some topics that interest me a lot: tail latency and the systems behind the modern internet.

Today, G5 Labs is emerging from stealth with $14 million in seed funding to build a new abstraction layer for AI-native software development. @pillar_vc and @BatteryVentures co-led the round, with participation from @omegavp , Encoded Ventures, and prominent angel investors including @JeffDean . G5 Labs is pioneering the post-developer era. A spinout from MIT’s Computer Science and Artificial Intelligence Laboratory (CSAIL), the G5 Labs platform makes natural-language intent, organized as an ontology, the source code itself—putting software development in the hands of the whole organization while providing the governance needed to keep requirements, implementation, and organizational policies aligned. We’re excited to build a future where employees across departments, from product owners to analysts and compliance leads, can create and evolve software using the language they already speak. Read our full announcement: Read more in VentureBeat:

Introducing OM-1, our first robot foundation model, zero-shot generalizing to any robot: table-top arms, industrial arms and humanoids. - learned directly from human manipulation data - no teleop/robot data - close to human-level dexterity and efficiency - multi-robot collab