GPT reasoning models have "line of sight" to AGI, says OpenAI's Greg Brockman

THE DECODER / 4/2/2026

💬 OpinionSignals & Early TrendsIdeas & Deep AnalysisModels & Research

共有:

Key Points

OpenAI co-founder Greg Brockman argues that the question of whether text-based models can reach general intelligence is effectively settled in favor of achieving AGI.
He claims GPT architectures—specifically reasoning-oriented GPT models—provide a “line of sight” to AGI.
The article frames Brockman’s stance as a direct rebuttal to skepticism about scaling from language models to general intelligence.
It emphasizes that, in his view, the path from current GPT reasoning capabilities to AGI is straightforward rather than uncertain.

OpenAI co-founder Greg Brockman says the debate about whether text-based models can achieve general intelligence is settled. The GPT architecture will lead to AGI.

The article GPT reasoning models have "line of sight" to AGI, says OpenAI's Greg Brockman appeared first on The Decoder.