Welcome to Science Reader!

What we do here
Saturday , 18 July 2026

Welcome to Science Reader!

What we do here
Saturday , 18 July 2026
HomeAI benchmarks

AI benchmarks

3 Articles
ai solves paul erdos problem
AI and computer scienceScience Explorer

AI Solves Erdős Math Problem: What's Next for AI in Mathematics?

An AI solved an 80-year-old Erdős math problem by walking a path mathematicians had collectively avoided.

May 23, 2026
A robot taking a test with pen and pencil.
AI and computer scienceScience Explorer

Humanity's Last Exam: The Hardest AI Benchmark Yet

AI benchmarks are meant to measure what machines can and cannot do. But the machines keep outsmarting the tests. A new test tries...

March 24, 2026
A sheep walking in a pasture, observing a computer screen which shows that the computer sees it as a zebra.
AI and computer scienceScience Tracker

AI Hallucinations Are Inevitable, OpenAI Researchers Prove

OpenAI researchers prove AI hallucinations are mathematically inevitable. Nine out of ten benchmarks reward guessing over admitting uncertainty.

September 14, 2025

Featured story

ai in science connects the dots

AI In Science Connects the Dots, But Only In Fields That Are Fragmented

July 3, 2026

What's new

the pope save humanity from ai

"Keep Humanity Safe From AI," Urges Pope Leo XIV

June 9, 2026
A woman shouting at her monitor, arguing with her AI chatbot.

Is AI Making You Dumber? Not If You Challenge It

April 20, 2026
Editorial illustration showing the Orion spacecraft about to reenter the atmosphere.

Artemis II Flew on AI, but Came Home on Engineering

April 15, 2026
An illustration showing a comet with a tail.

3I/ATLAS: The Interstellar Comet That Defied Expectations

April 8, 2026