Stephanie Zhan
@stephzhan
GP @Sequoia, early stage AI. On superintelligence quest. 1st partner to @linear @heytavus @mach_industries @middeskhq @replicatehq @reflection_ai @skildai
🔴Excited to share my TED AI talk! How we advance AI into an era of superintelligence, with true *practical* impact in our world. ❓I ask: For all the groundbreaking milestones we’ve reached, all the intelligence benchmarks we’ve shattered, for the $53B invested in generative…
New job alert: Make Linear's identity more prismatic 💠 We are at an interesting inflection point. Large companies are waking up to the fact that quality tools actually matter. Enterprise is our fastest growing segment -- but we still have a long road ahead of us. Hundreds of…
episode drop! ex-deepmind researchers @MishaLaskin @real_ioannis announce @reflection_ai's 1st product, Asimov, a "code understanding agent." Topics: *engineering != writing code *future of ai code *team-wide ai memory *scaling deep RL *why ASI will get deployed vertically
Ex-DeepMind researcher @MishaLaskin on building superintelligence, coding as a shortcut, reinforcement learning, launching @reflection_ai's first product, Asimov, and competing with the big AI labs 00:00 - Intro 01:42 - Reflection AI: Company Origins and Mission 04:14 -…
A space should inspire people to do their life’s work this one did. come build the next one with us...ft. more phone booths
Neat example of AI Software. Hard to call this an “agent” or to define it purely in English. It’s rich software, with smart modular components and representations. Statistical models will not eat the world. Software will absorb models into Modules. Programming, not prompting.
Engineers spend 70% of their time understanding code, not writing it. That’s why we built Asimov at @reflection_ai. The best-in-class code research agent, built for teams and organizations.
I've been really excited about this team and this launch for a while now. Really looking forward to getting my hands on it!
Engineers spend 70% of their time understanding code, not writing it. That’s why we built Asimov at @reflection_ai. The best-in-class code research agent, built for teams and organizations.
Plenty of brittle + narrow tooling in this wild west era of codegen can be characterized—not unfairly, and with just a hint of sarcasm—as off-the-shelf LLMs whose untreated amnesia only worsens an already-acute addiction (kink?) to github READMEs. We're a far cry from Fred…
Engineers spend 70% of their time understanding code, not writing it. That’s why we built Asimov at @reflection_ai. The best-in-class code research agent, built for teams and organizations.
big day at @reflection_ai listen to @MishaLaskin talk all about the new eng team secret weapon on @tbpn !
Morning. On today’s stream: – Why Grok is going all-in on anime mode – AWS drops a Browserbase competitor – What AI product is Zuck actually building? – @zebulgar (Varda) on hosting Trump for an AI event – @LucasSwisher1 (Coatue) on the state of U.S. markets – @GuptaRK22…
Very cool release by Misha, Ioannis and team at Reflection! Can’t wait to try it.
Engineers spend 70% of their time understanding code, not writing it. That’s why we built Asimov at @reflection_ai. The best-in-class code research agent, built for teams and organizations.
First launch with the Reflection team! We built the agent we always wanted when the code made no sense and the comments lied.
🆕 Post-Training with RL for Autonomous Coding youtube.com/watch?v=QluDzK… Congrats to @achowdhery for launching Asimov today! Building highly capable coding agents is a crucial step toward superintelligence more broadly. Aakanksha spoke about why @reflection_ai is investing so…
really cool product launch by @MishaLaskin and the team at @reflection_ai! we're currently more bottlenecked by code understanding than code generation
Engineers spend 70% of their time understanding code, not writing it. That’s why we built Asimov at @reflection_ai. The best-in-class code research agent, built for teams and organizations.
Code understanding agents are kind of the opposite of vibe coding -- users actually end up with *more* intuition. It's also a problem ripe for multi-step RL. Contexts are huge, so there's large headroom for agents that "meta" learn to explore effectively. Exciting to push on!
Engineers spend 70% of their time understanding code, not writing it. That’s why we built Asimov at @reflection_ai. The best-in-class code research agent, built for teams and organizations.
Today we launch Asimov. Asimov is our code research agent that is best-in-class in codebase comprehension. It is built for teams, built for enterprises, and built to remember. We use it everyday to accelerate our velocity and streamline distributed ops. Link below to sign up…
Andrej recently pointed out that vibecoding speed can’t explode to infinity because humans take time to read code But what if code comprehension were accelerated 10x? Great to see a cracked ex-gdm/oai team working on this problem, congrats on the launch!
Engineers spend 70% of their time understanding code, not writing it. That’s why we built Asimov at @reflection_ai. The best-in-class code research agent, built for teams and organizations.
how far can vibe coding take us? what’s missing? i’d argue… → what’s missing is *code understanding* code understanding is the reason your team’s veteran engineer is far more productive than another similarly talented person off the street. code understanding is what enables…
Engineers spend 70% of their time understanding code, not writing it. That’s why we built Asimov at @reflection_ai. The best-in-class code research agent, built for teams and organizations.
Excited to see @reflection_ai ship Asimov, the code-research agent. Why it matters: - Engineers waste ~70% of their week re-reading code, not writing it. - Most AI tools chase code generation and choke on real-world repos. - Lost context = stalled roadmaps and runaway burn.…
🎯 @reflection_ai team has entered the chat! Punching with the most powerful strategy, at the level of the frontier labs. Our two core beliefs: - the most impactful problem is code comprehension, not just code generation - the most performant products are built by those who…
Engineers spend 70% of their time understanding code, not writing it. That’s why we built Asimov at @reflection_ai. The best-in-class code research agent, built for teams and organizations.
Kimi K2 just dropped, and it's not "yet another larger-than-last-week" LLM. It's the first production-grade agentic model—built to act, not merely think. Think of DeepSeek R1 as the contemplative chess grandmaster: brilliant at analysis, but someone else moves the pieces. K2 is…