PLAY PODCASTS
Did GPT-5 Lose OpenAI's Crown? | Competition, Spicy ML Models, Reviews, & AI News
Episode 117

Did GPT-5 Lose OpenAI's Crown? | Competition, Spicy ML Models, Reviews, & AI News

*OpenAI's Crown Is Falling 👑 Did GPT-5 just blow it?* In this episode of *They Might Be Self-Aware*, Hunter Powers and Daniel Bishop dive deep into the chaotic reception of GPT‑5 — OpenAI's latest (and supposedly greatest) model — and explore whether it's actually a downgrade wrapped in hype. Has GPT‑5 lost OpenAI its crown? We break down the slow rollout, controversial benchmarks, and user backlash over “terse” replies, hallucinations, and capacity cost-cutting. Is GPT‑5 really the next leap forward… or just cheaper to run? But that’s not all. We explore how competitors like Claude 3, Gemini, MidJourney, and Runway are quietly (or not-so-quietly) eating OpenAI’s lunch. From jaw-dropping video generation to one-million-token context windows, the AI race is heating up—and OpenAI might be trailing. We also tackle: * 🤖 ChatGPT “psychosis,” AI as a therapist/friend, and the dangers of artificial intimacy * 📉 A real case of GPT-powered health advice leading to 19th-century bromism * ⚖️ Whether AI companies should be legally liable for bad advice or hallucinations * 🔬 AI benchmarks and what they *really* mean — plus the odd graphs from OpenAI's demo * 🧠 The theory that we’ve hit the ceiling on LLMs… or is this just a compute bottleneck? This isn’t just an AI update — it’s an honest (and hilarious) look at the messy reality of where frontier models are headed, and what OpenAI’s moves say about the future of the industry. --- 🎧 *Listen & Subscribe* 📱 Listen on Spotify: https://open.spotify.com/show/3EcvzkWDRFwnmIXoh7S4Mb?si=3d0f8920382649cc 🍎 Subscribe on Apple Podcasts: https://podcasts.apple.com/us/podcast/they-might-be-self-aware/id1730993297 ▶️ Subscribe on YouTube: https://www.youtube.com/channel/UCy9DopLlG7IbOqV-WD25jcw?sub_confirmation=1 --- *#GPT5 #OpenAI #AInews #ArtificialIntelligence #ChatGPT #Claude3 #Gemini #TMBSA* --- Enjoyed this episode? Give us a thumbs up, leave a comment, or hit ⭐️ on your favorite platform. It helps more curious minds discover the show.

They Might Be Self-Aware · Daniel Bishop, Hunter Powers

August 22, 202534m 53s

Audio is streamed directly from the publisher (cdn.simplecast.com) as published in their RSS feed. Play Podcasts does not host this file. Rights-holders can request removal through the copyright & takedown page.

Show Notes

CHAPTERS
0:00 Intro – GPT‑5, Claude's Funeral, and AI Burnout
2:15 Why GPT‑5 Disappointed (and Ticked Off) Hunter
5:10 Terse Replies, Token Saving & Capacity Games
8:00 Slow, Hallucinating, and Losing Trust
10:00 Fake Graphs, Benchmark Spin, and “Routing”
13:40 Who’s Actually Winning? (Spoiler: Not OpenAI)
17:00 Asymptotic LLMs, and the Future of Intelligence
20:00 GPT‑5 vs Claude vs Gemini vs Sora
23:00 ChatGPT Psychosis, Artificial Friends, and Lawsuits
26:00 GPT‑5 Coding Failures & Looping Bugs
30:00 Outro – Are We Just Replacing People with Bots?

Topics

openai newsai hypegpt-5 breakdowngpt-5 reviewopenai lawsuitchatgpt poisoninggpt-5gpt vs geminigpt-5 slowai replacingai modelsgpt-5 leakai benchmarksopenai crown