According to a recent Forbes report, former xAI researcher Eric Zelikman is leading the formation of an AI startup, Humans&, which has already assembled a team of senior researchers from top institutions such as Google, Meta, Anthropic, OpenAI, and DeepMind.
The company plans to raise up to $1 billion at a valuation of approximately $5 billion, demonstrating the market's high recognition of its technological vision and development potential.
Humans& aims to build AI systems capable of remembering and accurately responding to users' personalized preferences. Its core innovation lies in proposing a new training paradigm that emphasizes "human-machine collaboration."
It is revealed that this technological approach is expected to consume significantly more computing resources than current mainstream methods, reflecting a significant leap in model complexity and data processing capabilities.
The co-founder lineup is impressive, including Georges Harik, an early core member of Google with a strong influence in Silicon Valley's startup and technology investment circles; Noah Goodman, Associate Professor of Computer Science at Stanford University, whose research in computational cognitive science and human-computer interaction provides crucial academic support for the project;
and Andi Peng, who previously led post-training and behavior reinforcement work at Anthropic, whose experience is expected to contribute to the safety and controllability of the model in practical applications.
Despite the team's impressive background, Humans& has not yet announced specific product plans or launch timelines.
This strategy leaves ample room for exploration in its technological development, leading to high expectations from the industry for its next move.
In today's increasingly competitive artificial intelligence landscape, whether Humans& can carve out a niche in a field dominated by giants with its "human-centered" technological philosophy and top-notch team has become a closely watched frontier issue in the tech world.