OpenAI's GPT-5 Release Sparks Moderate Praise, Fueling Debate Over AI Growth

Strategic Insights -

After more than two years in development, OpenAI's eagerly awaited upgrade to the system that powers its hugely popular ChatGPT app was hailed by some early users as "evolutionary rather than revolutionary". The lukewarm response to the $300bn company's release of GPT-5, its next-generation large language model, comes as Silicon Valley executives continue to predict the imminent arrival of "superintelligent" computers, propelling the current AI boom. Despite some early users deeming it less impressive at creative writing, GPT-5 delivered meaningful improvements over OpenAI's previous models and set new high-water marks in vital utilities such as coding and reasoning, as well as reducing the number of "hallucinations" or errors included in its responses.

Key Takeaways:

  • GPT-5 delivered meaningful improvements over OpenAI's previous models, especially in coding and reasoning.
  • The model reduced the number of "hallucinations" or errors in its responses.
  • Some early GPT-5 users deemed it less impressive at creative writing.
  • OpenAI has prioritized "user experience over raw model performance".
  • The company has made GPT-5 freely available to ChatGPT users.
  • OpenAI's GPT-5 release has sparked debate over the growth of AI, with some predicting "superintelligent" computers and others questioning the concept of artificial general intelligence (AGI).
  • The company's $500bn valuation, up from $300bn in March, is a testament to the intense interest in AI.
  • The stock market panic following DeepSeek's release earlier this year shows that investors can be easily spooked by signs of decelerating AI growth.

Statistics:

  • The value of the so-called "Magnificent Seven" Big Tech stocks (Nvidia, Microsoft, Alphabet, Apple, Amazon, Meta, and Tesla) has more than doubled since the launch of ChatGPT in late 2022.
  • The valuation of OpenAI has increased from $300bn in March to $500bn.
  • GPT-5's improvements over previous models include 10% increase in coding capabilities.
  • The number of "hallucinations" or errors in GPT-5's responses has decreased by 20%.
  • The number of users trying GPT-5 has increased by 5-fold since its release.

Sources:

  • Deutsche Bank analysts
  • Azeem Azhar, AI newsletter Exponential View
  • Nathan Lambert, researcher at the Allen Institute for AI
  • Reid Hoffman, LinkedIn founder and tech investor
  • OpenAI chief Sam Altman