David Silver, the mastermind behind Google’s AlphaGo that famously beat the world champion at Go, just launched a new AI company called NSHP. The startup raised $1 billion and claims current AI systems like ChatGPT are heading down the wrong path.
Silver thinks today’s AI is too focused on copying human behavior instead of actually learning and adapting. His new company wants to build what he calls “superlearners” – AI systems that can figure things out on their own, just like AlphaGo learned to play Go better than any human ever could.
The Chess Master’s Next Move
At Google DeepMind, Silver created AI that didn’t just memorize chess moves or Go strategies. Instead, it played millions of games against itself and discovered winning strategies that humans never thought of. Now he wants to apply that same approach to solve bigger problems beyond games.
NSHP isn’t sharing many details about what exactly these “superlearners” will do. But Silver believes the key is building AI that learns through trial and error, not just by reading text from the internet like ChatGPT does.
The company already has backing from major investors who clearly believe Silver can recreate his AlphaGo magic. If he’s right about current AI taking the wrong approach, his new systems could leapfrog everything we’re seeing from OpenAI and others today.
