Amazon is betting on agents to win the AI race
SMRTR summary
A quiet revolution is unfolding in artificial intelligence that goes far beyond chatbots. David Luan, head of Amazon's AGI research lab, believes AI agents will soon become the foundation of computing itself.
"Agents are going to be the atomic building blocks of computing," Luan explains from his San Francisco lab, where he leads a small, elite team working on AI systems that can reliably execute real-world tasks.
After helping develop GPT-2, GPT-3, and DALL-E at OpenAI, Luan sees today's AI landscape reaching a plateau of sorts. Top models like GPT-5 are converging in capabilities, making benchmarks increasingly irrelevant.
What excites him now is creating AI that learns through trial and error, not just pattern recognition. His team uses reinforcement learning and self-play techniques similar to those DeepMind employed to master Go.
The goal? AI that understands consequences, not just mimics text.
"What keeps me going is the sheer amount of leverage we could give to humans on their time," says Luan, defining his vision of AGI as "a model that could help a human do anything they want to do on a computer."
SMRTR provides this summary for quick context. The original article belongs to The Verge.
Read the original article