Everyday AI Podcast – An AI and ChatGPT Podcast

Ep 744: Leaks show insights on OpenAI and Anthropic’s next models, Claude Computer Use goes viral and more

March 30, 2026

AI Summary

5 min read

This episode recaps key AI developments from the week ending March 30, including leaks on powerful upcoming models from OpenAI and Anthropic, Anthropic's viral Claude Computer Use feature, Microsoft Copilot enhancements, and various enterprise tools, integrations, and research advances. Host Jordan Wilson highlights implications for productivity, enterprise shifts, and agentic AI while noting the blistering pace of releases.

Upcoming Frontier Models

OpenAI has completed pre-training on a new model codenamed SPUD, with a potential reveal in weeks. CEO Sam Altman described it as capable of accelerating the economy, supporting a pivot to enterprise tools and a "productivity super app" integrating ChatGPT, Codex, and the Atlas browser. This aligns with shelving consumer projects like Sora and delaying others to focus resources on core business, robotics, and integrated workflows. Specific capabilities remain undisclosed but emphasize productivity gains, possibly in multimodal understanding, reasoning, or tool use.

Continue reading the full summary in the app — free to try.

Read Full Summary →

Free • No credit card required

Listen to Audio Summary Open in App

What you'll learn

1 (00:31) **OpenAI's SPUD Model Leak** - Reports reveal OpenAI completed pre-training on codenamed SPUD, potentially GPT-5 or beyond, with launch imminent
2 (05:33) **Anthropic's Claude Computer Use** - Research preview lets Claude control Mac computers for tasks like app use, file handling, and browser actions
3 (10:20) **OpenAI Revives Codex Plugins** - New plugin support for Codex enables custom skills, pre-packaged scripts, and integrations like Google Drive editing
4 (14:56) **Apple Opens Siri to Third-Party AI** - iOS 27 will route Siri queries to chatbots like Gemini or Claude; WWDC to highlight AI focus
5 (20:35) **Judge Halts Anthropic Govt Ban** - Federal judge blocks Pentagon directives labeling Anthropic a supply chain risk, allowing Claude use
6 (24:04) **Google's TurboQuant Compression** - New method shrinks LLM key-value cache memory 6x to 3 bits without retraining, speeding inference
7 (27:54) **Anthropic Model Leak: Mythos & Capabera** - Leak exposes draft on Claude Mythos (most powerful yet) and larger Capabera tier above Opus

+ Full timestamped outline available in the app

Show Notes

OpenAI and Anthropic are reportedly releasing new models any day now. 😱

A U.S. judge essentially said, 'No' to the federal government's quest to label Anthropic a supply chain risk.

And Google may have created a method that will change AI forever.

(Oh, and for our livestream audience, we'll be able to break some pretty big AI news live)

Miss any of this?

Don't worry. We bring you the AI News That Matters each week.

Newsletter: Sign up for our free daily newsletter
More on this Episode: Episode Page
Join the discussion on LinkedIn: Thoughts on this? Join the convo on LinkedIn and connect with other AI leaders.

Upcoming Episodes: Check out the upcoming Everyday AI Livestream lineup
Website: YourEverydayAI.com
Email The Show: [email protected]
Connect with Jordan on LinkedIn

Topics Covered in This Episode:

OpenAI SPUD Model Leak Details
Anthropic Claude Computer Use Goes Viral
Codex Plugin Support and Integration
Apple Siri AI Chatbot Integration iOS 27
Anthropic vs. U.S. Government Legal Battle
Google TurboQuant Model Compression Method
Anthropic Claude Mythos and Capabara Leak
Microsoft Copilot CoWork Frontier Launch
Multimodal Model Counsel Feature in Copilot
Major AI Industry Weekly Feature Roundup

Timestamps:

00:00 OpenAI's new AI model Spud

05:02 Anthropic's new AI feature

08:29 Improving AI agent communication

10:50 OpenAI adds plugin support to Codex

15:07 Apple opening Siri to AI chatbots

19:39 Start Here podcast series overview

20:36 Judge blocks federal ban temporarily

26:15 Advances in local AI models

29:58 Early access to new Anthropic model

34:15 Multimodal AI review workflow

36:58 Comparing models and weekly roundup

41:45 Promoting the podcast and newsletter

42:24 Closing and newsletter sign-up

Keywords:

OpenAI, Anthropic, AI model leaks, Claude Mythos, Capabara, GPT-5, GPT-4 Pro, Spud model, Claude Computer Use, MacOS AI agent, C