For obvious reasons, biohacking can be a serious issue even for general AI implementations, but when it comes to building a model strictly for biology, only a select few researchers will get access.
The Rosalind model is based on the latest internal research from OpenAI, and outperforms GPT-5.4, sometimes massively, on chemistry, biochemistry, genetics and experimental design — the datasets it was trained on.
Keep a human in the loop
It will not replace human researchers, but can aid them in analyzing and synthesizing huge amounts of research — which is common, for example in genetics.
The key question for life sciences will be how it deals with hallucinations, which can critically skew any research or discovery, Ars Technica writes.
OpenAI says the model is not good enough yet for autonomous research, and will rely on a healthy dose of human oversight.
So far, no AI-discovered drugs have made it through clinical trials, Axios notes.
Read more: OpenAI’s announcement, Ars Technica, Axios and Reuters.