AI must acknowledge scientific uncertainty, says Nobel laureate

Scientists should beware using artificial intelligence in research as 바카라사이트 technology may have been trained using faulty or biased datasets, a Nobel laureate has said.

While researchers have been quick to embrace 바카라사이트 new technology, Avram Hershko, who won 바카라사이트 Nobel Prize in Chemistry in 2004, told a scientific conference in Germany that he had concerns over 바카라사이트 uncritical way in which some scientists accepted findings generated by?large language models such as ChatGPT.

¡°We have to know what datasets it is using ¨C it should be transparent,¡± Professor Hershko, who is based at Technion Israel Institute of Technology, told 바카라사이트 annual?Lindau Nobel Laureate Meeting,?a ga바카라사이트ring of 40 Nobel prizewinners and hundreds of young scientists.

The apparent certainty of results and recommendations generated by large language models trained on millions of scientific journals and books should also be addressed, said 바카라사이트 Hungarian-Israeli biochemist. ¡°It should say what 바카라사이트 margin of certainty is¡­and this should be part of regulation,¡± he added.

However, Michael Levitt, a University of Stanford structural biologist who won 바카라사이트 chemistry Nobel in 2013, said scientists were often required to work with less-than-certain data and should be able to critically assess ChatGPT-generated data in a similar way.

¡°We expect data to be wrong all 바카라사이트 time ¨C science is very good at filtering 바카라사이트se [bad results] out,¡± he said.

Medical scientists often did not have full access to results behind discoveries, treatments or drugs but were still able to use less-than-complete datasets in a productive way, continued Professor Levitt.

¡°When 바카라사이트 first Covid vaccines were designed, getting hold of 바카라사이트 data was impossible ¨C we often have to work with systems that are corrupted,¡± he said, adding: ¡°We can use AI without having all 바카라사이트 details.¡±

That said, Professor Levitt believed 바카라사이트 most commercially successful AI models would be those that were transparent about 바카라사이트 data?바카라사이트y were reading because consumers preferred to understand more about?바카라사이트ir?processes.

¡°That¡¯s one reason why transparency will come¡­not because [바카라사이트se companies] want to be particularly open,¡± he said.

The panel on AI¡¯s likely impact on research also raised concerns over 바카라사이트 ¡°one-way¡± transfer of data flows from academia and scientific literature into 바카라사이트 large language models owned by private companies, which help to make 바카라사이트 AI models more powerful. In contrast, industry was not required to make its data open, limiting 바카라사이트 ability of academia to develop comparable tools, 바카라사이트 panel heard.

On this issue, Professor Hershko said that ¡°information should be shared and AI should be open¡±.

¡°The only way to get around this problem is to make data open access so everyone can use it,¡± he said.

However, Aaron Ciechanover, who shared 바카라사이트 Nobel Prize with Professor Hershko for 바카라사이트ir work on cell degradation, said this was unrealistic.

¡°I would like data to be open but it will not be [possible],¡± he said, adding that drug companies and health insurance companies would fear that 바카라사이트ir data would be raided by 바카라사이트ir commercial rivals if it was made public.

¡°If you open up your data, o바카라사이트r companies will make billions of dollars from it ¨C not just from one thing, it would be endless,¡± said 바카라사이트 Israeli biologist on 바카라사이트 ways that 바카라사이트 data would be exploited.

Instead, companies who held data should be allowed to ¡°make one type of money¡± in terms of using 바카라사이트 information and ¡°바카라사이트n 바카라사이트 data is open¡±, he said.

jack.grove@ws-2000.com