OpenAI's chief scientist says AI is getting close to being as good as a human research intern (www.businessinsider.com)

OpenAI's Chief Scientist, Jakub Pachocki, announced significant strides towards creating AI systems that can perform at the level of research interns. During a recent podcast appearance, he indicated that advancements in coding, math, and physics indicate that AI is on track to autonomously handle complex tasks. OpenAI aims to develop an "AI research intern" by 2026 and a fully autonomous researcher by 2028, marking a critical evolution in the capabilities of AI in technical fields. Pachocki highlighted the rapid advancements in coding tools like Codex, which are already significantly impacting programming workflows. He also pointed to math benchmarks as a pivotal measure for enhancing model reasoning. The goal is to increase the autonomy of AI systems to tackle specific technical challenges while working longer stretches independently—an essential step toward fully autonomous research. However, he acknowledged that current systems are not yet capable of independently conducting advanced research or model improvements. This progression underscores the potential of AI to revolutionize research processes, although the path to full autonomy remains a challenging endeavor.