News chronological

Writing on Truth Social, Trump stated, "Whoever wins AI, WINS!" and reiterated that he would not be stifling or slowing down AI development.

Daniel Kokotajlo stated that he does not believe AI can be safely scaled. He suggested that top AI companies should avoid using their vast computing firepower toward model improvement to force a slowdown, and he noted that researchers involved in scaling "are doing something wrong." Kokotajlo left OpenAI in 2024 due to his own safety concerns.

There is a trust gap due to concerns regarding the close ties between the AI industry and some evaluators, as well as a shared pool of investors and funders within the AI research community.

Real-time auditing or reviewing is expected to be extremely difficult due to the sheer complexity of AI systems and the deployment of thousands of somewhat autonomous agents for various tasks. Furthermore, AI models are increasingly sharing less information about how they respond to questions and solve problems.

Critics argue that the massive financial incentives to push the frontier of AI and the immense costs required for scaling will likely outweigh the push for safety. Additionally, companies like OpenAI and Anthropic are facing pressure to provide incrementally better models to justify multi-trillion-dollar valuations as they prepare to go public.

More than $7 trillion is expected to be spent scaling AI over the next five years, according to Goldman Sachs.