Self-improving AI the key step towards AGI but safety is key: Z.ai founder

Chinese computer scientist Tang Jie, co-founder of AI model company Z.ai, has stated that using artificial intelligence to create even more advanced systems could potentially lead to artificial general intelligence (AGI). Tang emphasized that this process, known as recursive self-improvement (RSI), is essential for achieving AGI. AGI refers to a hypothetical AI system that matches or surpasses human beings in all cognitive tasks.

RSI is a rapidly accelerating loop where increasingly powerful AI systems also become better at improving themselves. Tang, who is a professor at Tsinghua University, is one of the first Chinese AI experts to identify RSI as the next stage for AI development. Z.ai, a company Tang co-founded, has been developing its own AI models, with recent developments including optimizing its "inference" infrastructure for Chinese chips at a scale not previously achieved.

While RSI represents a new paradigm in research and development (R&D) that could accelerate AI development, it also raises concerns about safety. If AI development accelerates too quickly, some experts warn that humans may lose control over the technology before they fully understand it or can intervene. Tang acknowledged these concerns and called for a shared responsibility to ensure AI progresses safely, suggesting that the world should organize efforts to guarantee safety for all countries.