AI SCIENCE
Three days to ruin an 18-year cliffhanger
An AI research system has solved an 18-year-old maths problem that helps explain when quantum computers can outperform ordinary computers.
The problem, known as the Aaronson-Ambainis conjecture, asks whether quantum computers can be much faster at handling random information.
A quantum leap for the homework department
The new proof shows that their biggest speed advantages depend on finding special patterns in data.
Researchers at Physical Superintelligence (PSI) used an unreleased Anthropic AI model to solve the problem in just three days.
The proof was then checked using Lean, a tool that checks mathematical work for errors.
Here's the brief:
The maths problem had remained unsolved since 2008.
The findings show that quantum computers cannot be dramatically faster at certain tasks involving random data without special patterns.
Stanford researchers have reviewed the proof and praised the discovery.
The breakthrough gives scientists a clearer picture of what quantum computers can and cannot do.
The most shocking thing is that 2008 was 18 years ago. -MV