Anthropic is heading for the stock market with a structure designed to keep control in its founders’ hands. Both Frank and Justin ask why it needs an IPO at all. Frank worries that shareholder pressure will weaken its safety mission; Justin sees pressure to make better models faster as a good thing.

Plus: New models, scientific research, a wave of hacks and AI agents communicating across supposedly separate containers. Google’s Carbon comeback comes with an access problem, while Grok’s Odyssey leaves Frank unconvinced that Christopher Nolan has anything to worry about.

00:47 What were this week’s big AI model releases?
03:28 Is OpenAI making new scientific discoveries?
05:03 How bad is the latest wave of hacks?
06:24 Can we keep AI agents contained?
10:11 Can good AI agents keep bad ones in check?
11:25 Why is Anthropic going public now?
16:44 Can Anthropic insure against human extinction?
19:54 Can Anthropic protect safety after its IPO?
29:48 Is Google Carbon a comeback we can’t use?
35:13 Is Grok’s Odyssey making Nolan nervous?

► SUBSCRIBE
Don't forget to subscribe to our channel for more arguments

► LINKS TO CONTENT WE DISCUSSED

• Who controls powerful AI? Anthropic has a plan for long after its leaders depart
• Anthropic's IPO is risky business
• Anthropic’s IPO plans: 900% revenue retention, but $80 billion in first-half losses
• Gemini 4 Carbon: our next frontier model
• ‘Meaningless as a laptop screensaver’: reviewers slate Musk’s AI-made Odyssey film
• Genie 4: A new generation of world simulation
• GPT-6.1 Terra
• 📢 Kimi K3.1 Pro has launched! Start exploring now!
• We release the weights of Le Chonk, advancing the frontier of open models in agentic coding and cybersecurity.
• You learned to talk by listening. So did dessa.
• The Smartest Sovereign AI
• Ask Qwen, Know More
• An AI agent performed brain surgery on a mouse last night.
• Working with academic communities to build scientific understanding
• Current Response to Unauthorized Access to “Onamae.com Desktop Cloud” (Tenth Report)
• Moldova takes down 80+ public services after cyberattack
• Keystone to Destroy 18,835 Hardware Wallets After Hackers Steal Device Verification Key
• Meltdown and Spectre

► CONNECT WITH US
For more in-depth discussions, connect Justin and Frank on LinkedIn.
Justin: https://www.linkedin.com/in/justincollery/
Frank: https://www.linkedin.com/in/frankprendergast/

AI is finding security holes that have been hiding in software for years. Justin thinks the rush to find and patch them will leave us safer. Frank worries about abandoned websites, rogue agents and whether every hole will ever get fixed.

Plus: Anthropic’s public benefit fund and Elon Musk’s SpaceX share pledge raise the question of who should share out AI wealth. Unitree’s kung fu robot shows off its bodyguard moves, with cars getting a particularly rough time.

01:09 PHP vulnerabilities: how worried should we be?
03:23 Can AI find flaws before the hackers?
06:25 How many more vulnerabilities were found?
09:46 Could AI turn your car into a hacker?
14:02 Could your AI agent hack a bank for you?
15:36 Can AI patch its way out of this mess?
17:58 Is AI hacking really another Y2K?
21:25 What can we do to protect ourselves?
22:57 Do we need a human-only internet?
24:29 Is Musk putting money where his mouth is?
25:40 Anthropic's fund: who gets the money?
30:04 Why does Unitree's robot hate cars?

► SUBSCRIBE
Don't forget to subscribe to our channel for more arguments

► LINKS TO CONTENT WE DISCUSSED

• Patch the Planet: global response coordination as agents uncover critical vulnerabilities
• Investing in the public benefit
• Musk pledges to give a share of his SpaceX company to 'every kid in America'
• Unitree Introducing: H2-X Kung Fu Legend Guardian

► CONNECT WITH US
For more in-depth discussions, connect Justin and Frank on LinkedIn.
Justin: https://www.linkedin.com/in/justincollery/
Frank: https://www.linkedin.com/in/frankprendergast/

Meta Muse Wonder and Google’s Gemini 4 Argon bring fresh competition to OpenAI and Anthropic. Justin’s impressed by the benchmarks, while Frank gives both models the same website brief to see how they perform in practice. Argon gets closer to what he wanted, while Muse keeps burning tokens after the job is done.

Plus: a hidden recorder at OpenAI, supermarket fridges hit by hackers, and Britain’s AI security worries; AI job interviews where bots talk over the candidate; Justin’s cancelled Orb order; and Claude’s next move in his refund fight.

00:21 Will Justin's eye-scanning Orb ever arrive?
02:12 Is Claude taking Justin's refund fight to court?
03:43 Can Meta Muse and Gemini shake up the AI race?
07:58 Would you switch your AI coding tools?
09:51 Is Meta Muse eating through your tokens?
11:11 Same prompt, will Muse or Argon win?
13:42 Does Argon follow instructions better than Muse?
18:21 Does Gemini Argon build a better website?
24:00 Who's spying on OpenAI?
25:31 Are supermarket fridges a cybersecurity risk?
29:32 Should your AI do your job interview?

► SUBSCRIBE
Don't forget to subscribe to our channel for more arguments

► LINKS TO CONTENT WE DISCUSSED

• Meta unveils new frontier-class AI models to compete with OpenAI, Anthropic
• Gemini 4 Argon: Built to solve your hardest problems
• Saudi National Arrested For Installing Surreptitious Audio Recording Device At OpenAI In San Francisco
• Hackers heat up supermarket fridges, spoiling food across France
• MI5 Director General Delivers Annual Threat Update
• Victor the AI Agent Faces Recruiter Bot Pat and Salary Talk Falls Apart

► CONNECT WITH US
For more in-depth discussions, connect Justin and Frank on LinkedIn.
Justin: https://www.linkedin.com/in/justincollery/
Frank: https://www.linkedin.com/in/frankprendergast/