On social media, OpenAI's AI reasoning expert Noam Brown expressed his concern over the current abundance of "vague hype" surrounding artificial intelligence. While he acknowledges that advancements in the AI field bring optimistic prospects, he also points out that there are still many research issues that need to be addressed, emphasizing that OpenAI has not yet achieved what is termed "superintelligence."
Brown's statement sharply contrasts with some recent views within OpenAI. In January, OpenAI researcher Stephen McAleer mentioned, "I somewhat miss the days when we didn't know how to create superintelligence." This suggests that OpenAI may have found a clear path towards artificial superintelligence (ASI).
Before joining OpenAI, Brown worked at Facebook's Artificial Intelligence Research (FAIR), where his research focused on developing AI systems capable of defeating human players in complex games like poker and diplomacy. His work on the poker AI "Libratus" explored the concept of "test-time compute," indicating that increasing AI computation time can improve game strategies.
Brown later introduced these ideas to OpenAI, applying them to the development of language models. The company's latest model is based on this concept, opening up new avenues for development by increasing "thinking time" rather than merely enhancing training performance. He believes this alternative expansion approach will foster the emergence of new AI capabilities, stating, "Expansion in this area is still in its early stages."
Brown's perspective serves as a reminder that while the prospects of artificial intelligence are exciting, we must not overlook the existing limitations and challenges in current technology development.
Key Points:
🌟 Noam Brown pointed out that the AI hype on social media is overly vague and emphasized that many research issues remain unresolved.
🤖 Brown's views contrast with other OpenAI researchers, stressing that the company has not yet achieved superintelligence.
🧠 Brown mentioned expanding AI capabilities through "thinking time," believing there is still significant room for growth in this field.