scripod.com

#225: GPT-5.6, ChatGPT Work, Enterprise Agents, AI 2040 & Apple Sues OpenAI

Shownote

OpenAI shipped GPT-5.6, ChatGPT Work, and GPT-Live this week, and the real story is the move from chat to computer-use agents. Paul and Mike break down what changed, then get into enterprise AI agents and the "reverse information paradox," the AI 2040 repo...

Highlights

This episode covers a major shift in AI from chat-based interactions to computer-use agents, highlighted by OpenAI's launch of GPT-5.6, ChatGPT Work, and GPT-Live. The hosts also discuss enterprise AI adoption challenges, a new AI safety law in Illinois, and a cheating scandal at Brown University.
02:52
48% somewhat worry about AI labs absorbing data.
28:47
It feels like AGI.
39:49
Companies risk giving away proprietary knowledge when using AI models.
46:43
International cooperation is key to safely achieving superintelligence.
59:18
Apple sued OpenAI for trade secret theft.
1:06:14
Fines are small for major companies.
1:10:00
Top grade was a C+
1:13:11
Scores dropped from 96 to 48 when exams were in-person.
1:17:44
Media literacy is essential against foreign influence.
1:23:46
A simple prompt produced high-quality outputs in seconds
1:29:40
Prepare for AI's economic transformation despite uncertainty.

Chapters

Intro
00:00
GPT-5.6, ChatGPT Work, and GPT-Live (Plus GPT-6 Rumors)
05:45
Advice on AI Agents in the Enterprise
31:30
AI 2040
46:39
Apple Sues OpenAI for Trade Secret Theft
59:17
Illinois Signs Nation-Leading AI Safety Law
1:06:14
AI Safety Index: Nobody Gets an A
1:09:59
The AI Cheating Scandal at Brown
1:13:11
China and Russia Stoke US Public Opinion on Data Centers
1:17:44
AI Use Case Spotlight
1:20:57
AI Product and Funding Updates
1:26:19

Transcript

Mike Kaput: If you showed someone a recording of this five years ago, they'd be like, we have AGI. It's not as smart as the models you're going to be interacting with on your computer, but you'd just be like, wait a second, you can have an actual real-time...