Robot Safety Commands: MIT Robot Widens One Warning to Electronics; GPT-4o Errs Up to 9 of 105 Tests
· techtimes.com · popularity 20/100 · 0 similar stories today · AI Policy, Safety & Ethics (Artificial Intelligence)
Read the original at techtimes.com →
Summary
Robot safety commands gain broader reach in MIT's FLEX, which uses GPT-4o to turn one typed laptop warning into avoidance of phones and other electronics. FLEX scored 0.99 normalized reward against 0.69 for add-only editing, but each language-model operation erred in 3.8% to 8.6% of isolated trials, and the closed-loop user study ran only in simulation.
Found in: Tech Times. Rating basis: single outlet; outlet tier 2 (unlisted); 1 feeds; feed position 3.