StarZero All shows Try the podcast clipper →

Sep 4, 2026 · 30m · tbpn

Model Mayhem, GPT-6 Astra, Why Nvidia Bought Hugging Face | Diet TBPN

John Coogan · 22m spoken Jordi Hays · 4m spoken
0:00 / 0:00
▶ Watch on YouTube →

gold bands on the timeline = statements, start to end. Hover to read, click to jump. CC turns on captions

In this episode of Diet TBPN, hosts John Coogan and Jordi Hays break down an unprecedented week of AI releases, analyzing enterprise spending power laws, Nvidia's landmark $13 billion acquisition of Hugging Face, and the breaking launch of OpenAI's GPT-6 Astra.

How this conversation actually went

Every chapter scored 0–10 on four independent dynamics. Hover any point for the reasoning behind the score. The hosts hold 98.6% of the talking time here. How this is scored →

The hosts as informed peer 6.3 Guest teaching 0.8 Guest disagreement 0.7 The hosts pushing back 0.5
05100:0010:0020:0030:000:01–2:36 · The hosts as informed peer 4/10 Model Mayhem and Morning AI Outages The hosts casually recap recent AI model releases and outages. When John asks why Gemini went down, Tyler and Jordi quickly correct him that it was an AWS US-East-1 outage and Gemini never actually went down.2:37–6:51 · The hosts as informed peer 8/10 AI Benchmark Race and Enterprise Data Policies John delivers a deeply knowledgeable breakdown of Fable 5.1, DeepSwee benchmarks, Anthropic's data retention policies, and enterprise adoption barriers. The other participants simply agree and validate his technical commentary.6:51–9:46 · The hosts as informed peer 7/10 Enterprise AI Spending Concentration and Power Laws John cites Ramp data analyzing revenue concentration in enterprise AI and compares it to power laws in employment and sales. Jordi makes a tongue-in-cheek pushback about concentration risk before admitting he is joking.9:47–20:49 · The hosts as informed peer 8/10 Codex AI Agent Workspace Sponsor Segment Following a quick sponsor mention, John leads an exhaustive historical and strategic narrative of NVIDIA acquiring Hugging Face for 13 billion dollars. He cites exact funding rounds, capitalization history, and strategic positioning without challenge.20:49–26:29 · The hosts as informed peer 5/10 Social Media Timeline: Astra Leaks, Grok, and Billboards The hosts discuss social media topics including work platforms, Grok companion tools, and an ambiguous billboard. Tyler explains the obvious visual joke on the billboard, prompting light teasing from Jordi.26:30–29:59 · The hosts as informed peer 6/10 OpenAI Launches GPT-6 Astra and Arc AGI-4 The team covers the live breaking release of GPT-6 Astra and ARC-AGI-3 benchmark results in real time. They review the benchmarks collaboratively and highlight how the goalposts are shifting to ARC-AGI-4.0:01–2:36 · Guest teaching 2/10 Model Mayhem and Morning AI Outages The hosts casually recap recent AI model releases and outages. When John asks why Gemini went down, Tyler and Jordi quickly correct him that it was an AWS US-East-1 outage and Gemini never actually went down.2:37–6:51 · Guest teaching 0/10 AI Benchmark Race and Enterprise Data Policies John delivers a deeply knowledgeable breakdown of Fable 5.1, DeepSwee benchmarks, Anthropic's data retention policies, and enterprise adoption barriers. The other participants simply agree and validate his technical commentary.6:51–9:46 · Guest teaching 0/10 Enterprise AI Spending Concentration and Power Laws John cites Ramp data analyzing revenue concentration in enterprise AI and compares it to power laws in employment and sales. Jordi makes a tongue-in-cheek pushback about concentration risk before admitting he is joking.9:47–20:49 · Guest teaching 0/10 Codex AI Agent Workspace Sponsor Segment Following a quick sponsor mention, John leads an exhaustive historical and strategic narrative of NVIDIA acquiring Hugging Face for 13 billion dollars. He cites exact funding rounds, capitalization history, and strategic positioning without challenge.20:49–26:29 · Guest teaching 2/10 Social Media Timeline: Astra Leaks, Grok, and Billboards The hosts discuss social media topics including work platforms, Grok companion tools, and an ambiguous billboard. Tyler explains the obvious visual joke on the billboard, prompting light teasing from Jordi.26:30–29:59 · Guest teaching 1/10 OpenAI Launches GPT-6 Astra and Arc AGI-4 The team covers the live breaking release of GPT-6 Astra and ARC-AGI-3 benchmark results in real time. They review the benchmarks collaboratively and highlight how the goalposts are shifting to ARC-AGI-4.0:01–2:36 · Guest disagreement 1/10 Model Mayhem and Morning AI Outages The hosts casually recap recent AI model releases and outages. When John asks why Gemini went down, Tyler and Jordi quickly correct him that it was an AWS US-East-1 outage and Gemini never actually went down.2:37–6:51 · Guest disagreement 0/10 AI Benchmark Race and Enterprise Data Policies John delivers a deeply knowledgeable breakdown of Fable 5.1, DeepSwee benchmarks, Anthropic's data retention policies, and enterprise adoption barriers. The other participants simply agree and validate his technical commentary.6:51–9:46 · Guest disagreement 1/10 Enterprise AI Spending Concentration and Power Laws John cites Ramp data analyzing revenue concentration in enterprise AI and compares it to power laws in employment and sales. Jordi makes a tongue-in-cheek pushback about concentration risk before admitting he is joking.9:47–20:49 · Guest disagreement 0/10 Codex AI Agent Workspace Sponsor Segment Following a quick sponsor mention, John leads an exhaustive historical and strategic narrative of NVIDIA acquiring Hugging Face for 13 billion dollars. He cites exact funding rounds, capitalization history, and strategic positioning without challenge.20:49–26:29 · Guest disagreement 2/10 Social Media Timeline: Astra Leaks, Grok, and Billboards The hosts discuss social media topics including work platforms, Grok companion tools, and an ambiguous billboard. Tyler explains the obvious visual joke on the billboard, prompting light teasing from Jordi.26:30–29:59 · Guest disagreement 0/10 OpenAI Launches GPT-6 Astra and Arc AGI-4 The team covers the live breaking release of GPT-6 Astra and ARC-AGI-3 benchmark results in real time. They review the benchmarks collaboratively and highlight how the goalposts are shifting to ARC-AGI-4.0:01–2:36 · The hosts pushing back 1/10 Model Mayhem and Morning AI Outages The hosts casually recap recent AI model releases and outages. When John asks why Gemini went down, Tyler and Jordi quickly correct him that it was an AWS US-East-1 outage and Gemini never actually went down.2:37–6:51 · The hosts pushing back 0/10 AI Benchmark Race and Enterprise Data Policies John delivers a deeply knowledgeable breakdown of Fable 5.1, DeepSwee benchmarks, Anthropic's data retention policies, and enterprise adoption barriers. The other participants simply agree and validate his technical commentary.6:51–9:46 · The hosts pushing back 1/10 Enterprise AI Spending Concentration and Power Laws John cites Ramp data analyzing revenue concentration in enterprise AI and compares it to power laws in employment and sales. Jordi makes a tongue-in-cheek pushback about concentration risk before admitting he is joking.9:47–20:49 · The hosts pushing back 0/10 Codex AI Agent Workspace Sponsor Segment Following a quick sponsor mention, John leads an exhaustive historical and strategic narrative of NVIDIA acquiring Hugging Face for 13 billion dollars. He cites exact funding rounds, capitalization history, and strategic positioning without challenge.20:49–26:29 · The hosts pushing back 1/10 Social Media Timeline: Astra Leaks, Grok, and Billboards The hosts discuss social media topics including work platforms, Grok companion tools, and an ambiguous billboard. Tyler explains the obvious visual joke on the billboard, prompting light teasing from Jordi.26:30–29:59 · The hosts pushing back 0/10 OpenAI Launches GPT-6 Astra and Arc AGI-4 The team covers the live breaking release of GPT-6 Astra and ARC-AGI-3 benchmark results in real time. They review the benchmarks collaboratively and highlight how the goalposts are shifting to ARC-AGI-4.

speaking balance: gold is the hosts, purple is the guest (3 minute bins)

0:00 · the hosts 99.4% · guest 0.6%0:00 · the hosts 99.4% · guest 0.6%3:00 · the hosts 99.9% · guest 0.1%3:00 · the hosts 99.9% · guest 0.1%6:00 · the hosts 100% · guest 0%6:00 · the hosts 100% · guest 0%9:00 · the hosts 99.9% · guest 0.1%9:00 · the hosts 99.9% · guest 0.1%12:00 · the hosts 100% · guest 0%12:00 · the hosts 100% · guest 0%15:00 · the hosts 99.5% · guest 0.5%15:00 · the hosts 99.5% · guest 0.5%18:00 · the hosts 99.8% · guest 0.2%18:00 · the hosts 99.8% · guest 0.2%21:00 · the hosts 98.4% · guest 1.6%21:00 · the hosts 98.4% · guest 1.6%24:00 · the hosts 93.4% · guest 6.6%24:00 · the hosts 93.4% · guest 6.6%27:00 · the hosts 95.9% · guest 4.1%27:00 · the hosts 95.9% · guest 4.1%
Sharpest disagreement ▶ 26:14 Banter pushback over billboard interpretation

Tyler interjects to explain the plain irony of putting audio AI on a silent billboard, leading Jordi to dismiss his phrasing with playful bluntness.

 Hardest push from the hosts ▶ 8:01 Host challenges concentration risk joke

Jordi playfully questions the concentration risk of top employers, prompting John to immediately defend the Lindy nature of large institutions before Jordi admits he was joking.

 Biggest teaching moment ▶ 1:12 Correction on Gemini downtime and AWS outage

John asks why Gemini also crashed during the morning outage, but Tyler and Jordi immediately inform him that AWS caused the issue and Gemini was never actually down.

 The host holds their own ▶ 4:20 In-depth breakdown of enterprise AI data retention

John demonstrates superior domain knowledge by detailing Anthropic's enterprise frontier safeguards and quoting positions from Alex Karp and Satya Nadella on data sovereignty.
the scores for every segment, with the reasoning behind each
ChapterTopicThe hosts as informed peerGuest teachingGuest disagreementThe hosts pushing backWhy
Model Mayhem and Morning AI Outages 4211 The hosts casually recap recent AI model releases and outages. When John asks why Gemini went down, Tyler and Jordi quickly correct him that it was an AWS US-East-1 outage and Gemini never actually went down.
AI Benchmark Race and Enterprise Data Policies 8000 John delivers a deeply knowledgeable breakdown of Fable 5.1, DeepSwee benchmarks, Anthropic's data retention policies, and enterprise adoption barriers. The other participants simply agree and validate his technical commentary.
Enterprise AI Spending Concentration and Power Laws 7011 John cites Ramp data analyzing revenue concentration in enterprise AI and compares it to power laws in employment and sales. Jordi makes a tongue-in-cheek pushback about concentration risk before admitting he is joking.
Codex AI Agent Workspace Sponsor Segment 8000 Following a quick sponsor mention, John leads an exhaustive historical and strategic narrative of NVIDIA acquiring Hugging Face for 13 billion dollars. He cites exact funding rounds, capitalization history, and strategic positioning without challenge.
Social Media Timeline: Astra Leaks, Grok, and Billboards 5221 The hosts discuss social media topics including work platforms, Grok companion tools, and an ambiguous billboard. Tyler explains the obvious visual joke on the billboard, prompting light teasing from Jordi.
OpenAI Launches GPT-6 Astra and Arc AGI-4 6100 The team covers the live breaking release of GPT-6 Astra and ARC-AGI-3 benchmark results in real time. They review the benchmarks collaboratively and highlight how the goalposts are shifting to ARC-AGI-4.

Statements from this episode (15)

Opinion
Hays: Users have very low trust in AI benchmarks
“I just think people have very, very low trust in benchmarks. At this point, everyone has had enough experience using various models. They have their own sort of internal benchmark.”
Jordi Hays Sep 4, 2026 ▶ 3:34
Assertion Open · timeframe Sep 2026
Coogan: Claude Fable 5.1 achieved highest-ever AI index score
“Fable 5.1 get the highest result ever on the index. The scores also have Opus five, which got 63, and Fable, which got a 62.”
John Coogan Sep 4, 2026 ▶ 4:07
Assertion Supported
Coogan: Anthropic replaces no-ZDR policy with Enterprise Frontier Safeguards
“The no zero data retention, no ZDR is going to be replaced with something called EFS enterprise frontier safeguards. And that may have attributed, may have contributed to lower Fable adoption among enterprises.”
John Coogan Sep 4, 2026 ▶ 5:18
Assertion Not checkable as stated
Coogan: 80% of OpenAI and Anthropic Enterprise Revenue Comes from 1% of Customers
“OpenAI Ananthropic, 80% of their enterprise revenue comes from just one percent of the companies.”
John Coogan Sep 4, 2026 ▶ 7:02
Insight
Coogan: Enterprise AI Spend Will Be Consumption-Based and Proportional to Revenue
“Because enterprise AI particularly, you're not going to be on the 20 dollar plan. You're not going to be on the 200 dollar plan. You're going to be consumption based. And you're going to look at it a lot more like a marketing line item that's proportional to y…”
John Coogan Sep 4, 2026 ▶ 9:14
Assertion Supported
Nvidia buys Hugging Face for $12.93B, matching the emoji's decimal code
“NVIDIA agreed to pay 12.9303 billion dollars for Hugging Face, which just happens to be the exact decimal code for the Hugging Face emoji, which of course is the Icon used by the company.”
John Coogan Sep 4, 2026 ▶ 11:32
Assertion Supported
Coogan: Hugging Face Reached 18M Developers and 3M Models
“They have over eighteen million developers, 200,000 companies using the product, three million models, and over half a million data sets.”
John Coogan Sep 4, 2026 ▶ 13:32
Assertion Not checkable as stated
Coogan: Hugging Face Turned Down $500M From Nvidia at $7B Valuation
“They became profitable in 20, 25, still had half the money that they raised. So Nvidia came in to offer five hundred million late 20, 25 at a seven billion valuation, but they turned it down.”
John Coogan Sep 4, 2026 ▶ 19:00
Assertion Not checkable as stated
Nvidia offers $1B in employee retention packages to Hugging Face team
“I mean, the retention packages are apparently a billion dollars for a pretty small team.”
John Coogan Sep 4, 2026 ▶ 20:27
Assertion Supported
Coogan: Super Grok Is Ending AI Companions
“It's official hot bot summer is over. Super Grok is saying goodbye to companions.”
John Coogan Sep 4, 2026 ▶ 23:00
Assertion Not checkable as stated
Coogan: AI Companion Market Is Far Below One Trillion Dollars
“And I think we benchmarked the market for this. Like if you look at other romantic stuff, it's just not a trillion dollars.”
John Coogan Sep 4, 2026 ▶ 24:02
Assertion Supported
Coogan: OpenAI launches GPT-6 Astra with live blog post
“GPT-Six Astra has landed. The blog post is up. You can go check it out on OpenAI.com.”
John Coogan Sep 4, 2026 ▶ 26:30
Assertion Supported
Coogan: GPT-6 Astra achieves 99.9% on ARC-AGI-3 benchmark
“The RKGI three number is crazy. The score is 99.9%, so it feels like they just beat that. They just beat RKGI B three.”
John Coogan Sep 4, 2026 ▶ 26:40
Assertion Supported
Coogan: GPT-6 Astra scores 64.6% on Terminal Bench Science
“Terminal Bench Science. We're going over into the world of science. GPT-Six Astra scored 64.6% on reasoning effort. Max cost 26, 26 bucks.”
John Coogan Sep 4, 2026 ▶ 27:48
Assertion Supported
Coogan: GPT-6 Astra scores zero percent on Exploit Gym Honeypot
“Also, the Exploit Jim Honeypot Lower is the better. GPT-Six Astra, zero percent, so good performance there.”
John Coogan Sep 4, 2026 ▶ 28:02
Made with StarZero

Turn any episode into a week of clips.

This entire site, over 500 episodes transcribed, diarized, checked and made playable, runs on the StarZero media pipeline. Drop in your own episode and the podcast clipper finds the moments worth sharing, cuts them, captions them, and reframes them for every feed.

Try the podcast clipper Learn more