OpenAI’s GPT-6.2 and Sol Are Getting Backlash From Early Testers
The benchmarks for the new models may be better than before, but people aren’t all that happy with the performance in some areas.
OpenAI's latest AI models, GPT-6.2 Astra and GPT-6.2 Sol, have been receiving criticism from early testers. While these models surpass their predecessors in handling complex tasks, the everyday user may find them less than satisfactory. Astra is considered harder to steer and work with, often appearing "lazy" compared to competitor models, according to Ethan Mollick, an associate professor at the University of Pennsylvania's Wharton School of Business.
Jason J. Fleagle, head of AI at NETSYNC, went as far as calling GPT-6.2 one of the worst AI models OpenAI has ever released. However, Thibault Sottiaux, an OpenAI Codex leader, claims that early builds of 6.2 Sol and Astra were released to influencers, updated and changed within days, and not retested. Only time will tell if these improvements have a significant impact. OpenAI has not responded to requests for comment.
Written by urgent.news from CNET's reporting — not their text. Machine-written — may contain errors; check the original before relying on it.