#OpenAI’s #GPT5 shows promise in accelerating #scientificresearch, particularly in biology, maths, and algorithmic decision-making. While it can identify existing conclusions and even contribute to new discoveries, it still requires #humanoversight due to limitations like potential #hallucinations and sensitivity to scaffolding. OpenAI emphasises a #partneredapproach, where AI tools like GPT-5 enhance human expertise. https://www.zdnet.com/article/gpt-5-is-speeding-up-scientific-research-but-still-cant-be-trusted-to-work-alone-openai-warns/?AIagents.at #AIagent #AI #ML #NLP #LLM #GenAI
GPT-5 is speeding up scientific research, but still can't be trusted to work alone, OpenAI warns

The model showed impressive strides - but the company isn't taking this as an indicator that AGI is near. Here's why.

ZDNET