CEOInterviews.AI
Start App
Try: AI strategy tariffs earnings guidance layoffs interest rates growth

Rayan Krishnan

Co-founder & CEO, Vals AI

Search every verified Rayan Krishnan interview, podcast appearance, and on-the-record quote — each transcript cross-checked by AI and human review to confirm speaker identity. Rayan Krishnan, co-founder and CEO of Vals AI, has recently appeared in multiple media outlets and events discussing his company's independent benchmarking of AI models. At the AI Engineer NYC 2026 conference in October, he spoke about the role of evaluations in codifying a company's competitive advantage and the importance of testing increasingly agentic systems. In interviews with Bloomberg Technology and The Information, Krishnan discussed Vals AI's recursive self-improvement (RSI) index, which tracks frontier models' progress toward autonomously building their successors. He stated that based on five proxy tasks covering pre-training, post-training, and harness engineering, his firm's projections suggest Anthropic could reach full RSI between July and August 2027, meaning the work of machine learning engineers at the lab would be automated. Krishnan noted that models currently show early signs of RSI but remain well behind top human researchers. Krishnan has also commented on the broader AI safety debate, responding to calls from Anthropic CEO Dario Amodei and others to slow frontier model development. He said that while some statements have caused undue public anxiety, he sees reason for optimism based on engagement with major labs, members of Congress, and government agencies. He argued that investment in testing and evaluation has not kept pace with model capabilities, and emphasized the importance of independent evaluators that do not sell training data or remediation services to the labs they assess. On a16z's podcast, Krishnan discussed why self-reported model scores can be misleading, citing an example where Meta's Llama 4 underperformed on Vals AI's private benchmarks despite strong public benchmark results, and described the need for continuously evolving evaluations as models improve.

Recent appearances

  • AI Engineer NYC Livestream | Bridgewater, Two Sigma, Arize, Coatue, Nous (Hermes), Gen. Intuition

    October 13, 2026 · 📺 AI Engineer

    Schedule: https://ai.engineer/schedule Live from the Sheraton New York Times Square Hotel in New York City, AI Engineer New York 2026 is an AI conference for engineers and technical leaders building production AI across banking, hedge funds, fintech, insurance, and asset management. Event: AI Engineer NYC Dates: October 12–14, 2026 Venue: Sheraton New York Times Square Hotel Learn more about the event: https://ai.engineer/nyc/2026# Subscribe for more AI Engineer talks, livestreams, and event coverage: / @aidotengineer Chapters: 0:00 Stream begins 0:22 Opening Remarks — Lia McBride…

    Watch on YouTube ↗

  • Anthropic Will Reach Full RSI by August 2027 | Rayan Krishnan

    October 03, 2026 · 📺 MTS

    Vals AI CEO Rayan Krishnan discusses his company's new RSI Index benchmark, predicting Anthropic could reach full recursive self-improvement by summer 2027. The conversation covers the mechanics of automating AI research, Washington DC's changing regulatory landscape, and how AI automation will diffuse throughout the broader economy. Check out our sponsor Lovable: ⁠https://lovable.dev.⁠ Turn ideas into software people love. -- ⁠https://x.com/RayanKrishnan⁠ -- MTS is a live news and interview show covering technology, business, politics, and culture as it happens. Follow MTS:  X |⁠ ⁠⁠⁠  / mtsl…

    Watch on YouTube ↗

  • Evaluating OpenAI and Anthropic's Latest Models

    September 29, 2026 · 📺 The Information

    Vals AI Co-Founder and CEO Rayan Krishnan explains why his firm’s recursive self-improvement research currently points to Anthropic reaching a key human-researcher threshold before OpenAI if recent progress continues. He discusses what happens when models become capable of doing more of the work required to develop their successors. Subscribe: https://www.theinformation.com/subscr... The Information’s TITV airs weekdays on YouTube, X and LinkedIn at 10AM PT / 1PM ET. Or check us out wherever you get your podcasts. Follow us: X: https://x.com/theinformation IG:   / theinformation   TikTok…

    Watch on YouTube ↗

  • AI’s Frontier Faces Calls for Safety Slowdown

    September 14, 2026 · 📺 Bloomberg Tech

    Bloomberg’s Ed Ludlow breaks down calls from AI for a slower pace at the frontier, arguing safety needs time to catch up with increasingly powerful models. The debate is reaching Washington and Wall Street, with AI-linked stocks under pressure as investors weigh what slower development means for the spending boom. 00:00:00 - Bloomberg Tech Begins 00:01:11 - Rachel Metz & Mike Shepard, Bloomberg News 00:05:35 - Uday Cheruvu, Harding Loevner 00:12:56 - Rayan Krishnan, VALS AI 00:18:48 - Peter Elstrom, Bloomberg News 00:23:34 - Parmy Olson, Bloomberg Opinion 00:31:02 - David Baszucki, Roblox CEO…

    Watch on YouTube ↗

  • Can Independent Testing Make AI Safer?

    September 14, 2026 · 📺 Bloomberg Tech

    Vals AI co-founder and CEO Rayan Krishnan says investment in testing and evaluating AI hasn’t kept pace with the rapid improvement in model capabilities. His firm independently evaluates models from companies including OpenAI and Anthropic, and sees early signs of AI moving toward recursive self-improvement, though models remain well behind top human researchers today. He joins Ed Ludlow on "Bloomberg Tech." -------- Like this video? Subscribe to Bloomberg Technology on YouTube:    / @bloombergtech     Watch the latest full episodes of "Bloomberg Technology" with Caroline Hyde and Ed Ludl…

    Watch on YouTube ↗

  • Inside the Race to Measure Frontier Intelligence

    September 09, 2026 · 📺 a16z

    a16z’s Erik Torenberg, Ben Horowitz, and Jennifer Li sit down with Vals founder and CEO Rayan Krishnan to discuss one of AI’s increasingly difficult problems: how do you actually measure whether a model is getting better? As public benchmarks saturate and models get better at optimizing for the tests themselves, Rayan makes the case for independent, continuously evolving evaluations. They unpack why self-reported model scores can be misleading, how VALS evaluates models in the hours before a release, and why measuring increasingly agentic systems means testing work that can unfold over hours,…

    Watch on YouTube ↗

  • Rayan Krishnan | Unlocking AI's Potential Through Benchmarking

    February 03, 2026 · 📺 The Token Economy Podcast

    Watch on YouTube ↗

  • Harvard Medical AI: Rayan Krishnan on VOS: Learning What You Don't Know by Virtual Outlier Synthesis

    June 22, 2022 · 📺 Harvard Medical AI | Rajpurkar Lab

    A talk hosted by the Rajpurkar Lab at Harvard which works on developing medical AI. These talks cover recent papers or topics in core AI / medical AI in a format targeted to those interested in the cutting edge of AI and its applications in medicine. In this session, lab member Rayan Krishnan presents "VOS: Learning What You Don't Know by Virtual Outlier Synthesis" by University of Wisconsin. Du, Xuefeng and Wang, Zhaoning and Cai, Mu and Li, Yixuan (2022). VOS: Learning What You Don't Know by Virtual Outlier Synthesis. arXiv preprint arXiv:2202.01197. Full Paper: https://arxiv.org/abs/2202…

    Watch on YouTube ↗

The summary and quote tags on this profile are produced with AI assistance from verified, first-person interview transcripts, then checked by our team to confirm the speaker's identity and the accuracy of every quote. See how we verify →