Summary
GPT-5 was released with high expectations but disappointed many users and showed few improvements over earlier models. New research confirms that large language models like GPT-5 still struggle to generalize and reason beyond their training data. This suggests that simply making models bigger won’t bring us true artificial general intelligence (AGI).
Highlights
id932990968
Ultimately, the idea that scaling alone might get us to AGI is a hypothesis. No hypothesis has ever been given more benefit of the doubt, nor more funding. After half a trillion dollars in that direction, it is obviously time to move on. The disappointing performance of GPT-5 should make that enormously clear. Pure scaling simply isn’t the path to AGI. It turns out that attention, the key component in LLMs, and the focus of the justly famous Transformer paper, is not fact “all you need”.