OpenAI claims we've entered the AGI era — has GPT-6 Astra really demonstrated general intelligence?
OpenAI says its newly launched GPT-6 Astra model brings humanity to the threshold of AGI, but external evaluators argue that inflated lab metrics and opaque testing obscure its true reliability.
OpenAI claims that its latest model, GPT-6 Astra, represents the beginning of the artificial general intelligence (AGI) era. However, skeptics argue that the model's impressive performance in benchmarks relies on heavy prompt optimization rather than true general intelligence. OpenAI President Greg Brockman stated that Astra signaled a fundamental turning point in AI evolution, but independent researchers caution that benchmark results do not necessarily indicate the achievement of AGI.
While Astra scored highly in various tests, including software engineering, mathematics, and scientific research, experts warn that these results should not be taken as conclusive proof of AGI. The company's decision to abandon the planned rollout of GPT-6.1 Astra due to safety concerns adds to the skepticism surrounding the claim.
Written by urgent.news from Live Science's reporting — not their text. Machine-written — may contain errors; check the original before relying on it.
