🧀 BigCheese.ai

Social

GPT-4.5: "Not a frontier model"?

🧀

GPT-4.5, described as 'not a frontier model' by OpenAI, is nonetheless the largest model tested by the general public, raising questions but not signaling the end of the AI bubble. It exhibits reduced hallucinations and improved emotional intelligence but doesn't signify a large capability leap. Experts debate its nuances and OpenAI may not maintain it in the API, considering its high costs relative to efficiency improvements in the industry.

  • GPT-4.5 described as not a frontier model
  • The biggest model for the general public
  • Lacks a clear jump in capabilities
  • Raised questions about OpenAI's release strategy
  • High costs with marginal improvements