No Priors AI New Updates to GPT-5: The Ethics of Progress
14 snips
Oct 28, 2025 The discussion dives deep into the ethical implications of the latest updates to GPT-5. You'll learn about the restored model picker and new modes like Auto, Fast, and Thinking. Concerns over the cost of legacy models like GPT-4.0 and the popularity of the paid 4.5 are addressed. The potential emotional impact of nuanced model personalities on users, especially in therapeutic contexts, raises intriguing questions. Plus, there's a comparison with other AI platforms facing similar challenges, making this a thought-provoking conversation.
AI Snips
Chapters
Transcript
Episode notes
Model Picker Returns With Mode Choices
- OpenAI reintroduced a model picker with Auto, Fast, Thinking, and Pro to give users choice over response style and quality.
- The router that selects models was broken for GPT-5, prompting manual mode options and legacy model returns.
Choose Mode Based On Task Needs
- Use Auto for most queries since it routes to the best model automatically and is fixed now according to Sam Altman.
- Switch to Fast for quick answers or Thinking for deeper responses when you need higher-quality reasoning.
Legacy Models Restored For User Preferences
- OpenAI restored GPT-4o and legacy models after user backlash because some users preferred older personalities like GPT-4o's warmth.
- Paid Pro access now gates experimental high-compute models like 4.5 and Thinking Mini.