Researchers gave top artificial intellect models a classic attention test used in psychology and found a major flaw. Some leading systems fell from over 90% accuracy to nearly complete failure. While the models could correctly name colors in short lists, their performance deteriorated sharply as the task became longer and more complex.
Related Posts
Does ‘federated unlearning’ in AI improve data privacy, or create a new cybersecurity risk?
As governments push for stronger data rights like the “right to be forgotten,” evidence suggests artificial intelligence may not fully…
This new chip could slash data center energy waste
A fresh chip design from UC San Diego could make data centers far more energy-efficient by rethinking how power is…
Think AI "knows" what it’s doing? Scientists say think again
When they do, it often falls on a spectrum—sometimes describing simple requirements, other times hinting at human traits. Calling artificial…