OpenAI says it disproved a long-held idea about hard computer problems

OpenAI says it has disproved a long-held idea in computer science: that some problems will always be hard for computers to solve quickly. Russell Impagliazzo, who helped propose that idea, called the result "shocking news." The report comes from Cade Metz at the New York Times.

What changed

The idea is called the Exponential Time Hypothesis. It held that some problems would always be hard for computers to solve quickly. Computer scientists have used that belief as a working assumption for years. OpenAI now says it has shown the belief is wrong.

Why it matters

The finding could one day speed up chip design. It could also help break encryption, the scrambling that protects data and payments. Many security tools rely on certain problems being too hard to crack in a reasonable time. If those problems turn out to be easier, the locks could become easier to pick.

What is not known yet

• Whether the result holds up once other experts check it.
• How soon, if ever, it leads to faster chip design.
• Whether it affects the encryption used today, or only in theory.
• What the proof looks like in detail.

The source material gives only a short summary, and the full article was not available. The wording is cautious: the result "could one day" have these effects. Nothing here suggests that your bank or phone is at risk right now.

What happens next

Expect other researchers to examine the claim closely. Their verdict will show how much weight the result deserves and how far it reaches.