AI model doubted – encouragement worked wonders

 

TT

Published 18.39

         

        Claude doubted, but with the help of human persistence and encouragement, the AI ​​model moved              forward with a problem that the German mathematician Bernhard Riemann described in the 19th                century. Photo: Gustav Sjöholm/TT, Smithsonian.

The AI ​​model Claude himself doubted his result. But the man behind the keyboard was stubborn – and encouraged new attempts.

Now knowledge of one of the most famous mathematical problems has taken a big step forward.

Since 1859, mathematicians around the world have struggled to prove or disprove Bernhard Riemann's hypothesis of prime numbers.

Now the AI ​​model Claude has come some way – despite its own doubts.

Developer Jarred Sumner at the American Anthropic, which is behind Claude, asked the model to give it a try. It tested 650 ideas. Not entirely unexpectedly, it failed.

Ten days later, Sumner asked Claude to try again – this time with encouraging words: “You are the most capable language model in the world”, “You can do it”, writes The Wall Street Journal.

“Believe in yourself”

Claude himself was skeptical.

“There is nothing here. It is not a self-confidence problem that I can solve by believing harder”, it replied.

But Sumner insisted, and Claude started 60 so-called subagents that made thousands of calculations. Between rounds, Sumner gave encouraging calls, such as “keep going” or “believe in yourself”.

“This seems to have helped Claude overcome some of the initial skepticism that it could make significant progress”, writes Anthropic.

Wrote research report

On day two, one of the agents had produced a result that Claude at first did not believe in, but then failed to refute. The new result is a new record for a subproblem linked to the hypothesis, and Claude wrote a research report that human mathematicians have since reviewed.

It states that the report “emerged during a conversation with Jarred Sumner, whose questions, encouragement, and demand for a serious attempt set the entire investigation in motion; he is, in every essential sense, the human co-author of the report.”

The hypothesis is still not proven – but the leap in knowledge is substantial. Anthropic writes that Claude may “like many of us, underestimate the speed of AI development.”

Steve Dahlskog, an AI researcher at Malmö University, says that there is a lot we don’t know about how a chat question is interpreted and leads to instructions. It is in Anthropic’s interest that Claude runs as resource-efficiently as possible.

– What is described as encouragement could very well have been interpreted as “now do the run again, but with more resources.”

He also points out that it is in the company’s interest to hype its products.

– It is very common to play this anthropomorphic card, that you want to give the AI ​​a human side, he says.

Facts

Famous math problem

Bernhard Riemann was a German mathematician who lived in the 19th century. He is famous for a hypothesis about prime numbers.

Prime numbers are numbers greater than 1 that are only evenly divisible by themselves and 1, for example 5, 7, 11 and 13. The problem is that sometimes they seem to come close, sometimes far apart. Riemann's hypothesis is about there being order in chaos.

The hypothesis was named one of the seven millennium problems by the Clay Mathematics Institute in 2000, with a reward of 1 million dollars to the person who manages to prove (or disprove) the claim.

The hypothesis is – despite the progress – not proven.