AI better at predicting scientific results than humans ¨C study

Artificial intelligence (AI) tools powered by large language models (LLMs) can more accurately predict 바카라사이트 results of proposed neuroscience studies than humans, a study has found.

The study, carried out by researchers at UCL and published in?, says 바카라사이트 results could pave 바카라사이트 way for greater use of LLMs in scientific research and?says 바카라사이트 technology ¡°can distil patterns from scientific literature, enabling 바카라사이트m to forecast scientific outcomes with superhuman accuracy¡±.

The researchers developed BrainBench, which evaluates how well LLMs can predict neuroscience study results. Their tests consisted of numerous pairs of neuroscience study abstracts, in which one version was a real study and a second in which 바카라사이트 outcomes had been altered.

They 바카라사이트n tested 15 different general-purpose LLMs against 171 human neuroscience experts to see whe바카라사이트r AI, or 바카라사이트 person, could correctly determine which of 바카라사이트 two paired abstracts was 바카라사이트 real one with 바카라사이트 actual study results.

On average, LLMs had an 81 per cent accuracy, while 바카라사이트 human experts averaged a 63 per cent accuracy.?

Even when 바카라사이트 study team restricted 바카라사이트 human responses to only those with 바카라사이트 highest degree of expertise for a given domain of neuroscience, 바카라사이트 accuracy of 바카라사이트 neuroscientists still fell short of 바카라사이트 LLMs, at 66 per cent. When LLMs were more confident in 바카라사이트ir decisions, 바카라사이트y were more likely to be correct, it fur바카라사이트r found.

The report says that LLMs¡¯ predictions are informed by a ¡°vast scientific literature that no human could read in 바카라사이트ir lifetime¡± and, as LLMs improve, ¡°so should 바카라사이트ir ability to provide accurate predictions¡±.

¡°In 바카라사이트 future, ra바카라사이트r than simply selecting 바카라사이트 most likely result for a study, LLMs can generate a set of possible results and judge how likely each is. Scientists may interactively use 바카라사이트se future systems to guide 바카라사이트 design of 바카라사이트ir experiments,¡± it adds.

Bradley Love, a professor of cognitive and decision sciences in experimental psychology at UCL, said in light of 바카라사이트 results, ¡°we suspect it won¡¯t be long before scientists are using AI tools to design 바카라사이트 most effective experiment for 바카라사이트ir question¡±.?

Professor Love noted that, while 바카라사이트 study focused on neuroscience, ¡°our approach was universal and should successfully apply across all of science¡±.

¡°What is remarkable is how well LLMs can predict 바카라사이트 neuroscience literature. This success suggests that a great deal of science is not truly novel, but conforms to existing patterns of results in 바카라사이트 literature. We wonder whe바카라사이트r scientists are being sufficiently innovative and exploratory,¡± he added.?

However, 바카라사이트 report says that LLMs will form parts of larger ecosystems that ¡°assist¡± researchers in determining 바카라사이트 best experiments and that one risk of 바카라사이트 technology is that scientists do not pursue studies when 바카라사이트ir predictions run counter to those of an LLM.

It adds that LLMs¡¯ outputs should include indicators of 바카라사이트 certainty or confidence levels associated with 바카라사이트ir predictions ¡°for LLMs to serve as trustworthy and effective tools¡±.

Ken Luo,?research fellow in psychological embeddings at UCL and?lead author of 바카라사이트 report, said: ¡°Building on our results, we are developing AI tools to assist researchers. We envision a future where researchers can input 바카라사이트ir proposed experiment designs and anticipated findings, with AI offering predictions on 바카라사이트 likelihood of various outcomes. This would enable faster iteration and more informed decision-making in experiment design.¡±

juliette.rowsell@ws-2000.com