Ex-DeepMind Scientists Want Their AI to Be Smarter Than You

Remember when AI was just helping coders catch typos? Those were simpler times. Now, two DeepMind veterans have raised $130 million to build AI that won't just assist developers - it'll replace them entirely. And that's just the beginning.

Reflection AI, valued at $555 million before even emerging from stealth, wants to create superintelligent AI agents that can code autonomously. Think less "helpful assistant" and more "your new robot overlord." Founders Misha Laskin and Ioannis Antonoglou bring serious credentials - they helped create AlphaGo and Gemini at DeepMind. Now they're aiming higher.

Big Money Backs the Robot Revolution

The startup has already attracted big names. Reid Hoffman, Nvidia, and Sequoia Capital are all throwing money at the chance to help build potentially humanity-surpassing AI. Current clients include financial firms and tech companies, who are apparently eager to hand their codebases over to autonomous agents.

Humans Need Not Apply (Much)

While other companies build AI "co-pilots," Reflection wants their AI in the captain's chair. They're starting with automating tedious coding tasks, but their ambitions stretch far beyond that. As one investor helpfully clarified, they don't want to replace engineers - they just want to turn them into "architects" overseeing armies of AI agents. Much better.

The timing seems perfect, riding the wave of AI enthusiasm that followed ChatGPT's debut. With offices in New York, San Francisco, and London, Reflection is now hiring. Presumably, they'll need humans. For now.

Why this matters:

  • We're witnessing the tech industry's shift from "AI will help humans work better" to "actually, do we need humans at all?"
  • When the people who built some of the world's most advanced AI systems leave to pursue "practical superintelligence," it might be time to update your resume. Or learn to welcome our new AI overlords.

Read on, my dear: