The Don’t Worry About the Vase Podcast is a listener-supported podcast. To receive new posts and support the cost of creation, consider becoming a free or paid subscriber. * 00:00:00 - Introduction * 00:03:40 - Table of Contents * 00:07:24 - Language Models Offer Mundane Utility * 00:11:03 - Language Models Don’t Offer Mundane Utility * 00:14:37 - Levels of Friction * 00:16:51 - Huh, Upgrades * 00:17:09 - On Your Marks * 00:17:22 - Lack of Cybersecurity * 00:19:03 - Meta Game * 00:26:24 - Deepfaketown and Botpocalypse Soon * 00:27:18 - A Young Lady’s Illustrated Primer * 00:29:01 - Let My People Go * 00:30:15 - You Drive Me Crazy * 00:30:56 - They Took Our Jobs * 00:35:28 - They Gave Us Time Off * 00:41:42 - Get Involved * 00:42:52 - Introducing * 00:43:16 - In Other AI News * 00:48:33 - Thanks For The Memos * 00:51:27 - Show Me the Money * 00:53:54 - Bubble, Bubble, Toil and Trouble * 00:54:45 - Quickly, There’s No Time * 00:55:28 - The Quest for Sane Regulations * 00:57:59 - Our Offer Is Nothing * 01:03:12 - The Week in Audio * 01:03:30 - Rhetorical Innovation * 01:09:50 - Political Violence Is Never The Answer * 01:13:23 - A Lot Of People Peacefully Speak Of Infinitely High Stakes * 01:15:01 - Take a Moment * 01:19:08 - Greetings From The Department of War * 01:24:34 - Political Pressure At Google DeepMind * 01:25:10 - Things That Are Basically Legal And Accepted Now, Somehow * 01:26:12 - Aligning a Smarter Than Human Intelligence is Difficult * 01:32:11 - Aligning a Current Model For Mundane Tasks Is Also Difficult * 01:33:29 - Everyone Is Confused About AI Consciousness * 01:36:30 - The Lighter Side https://open.substack.com/pub/thezvi/p/ai-164-pre-opus?utm_campaign=post-expanded-share&amp;utm_medium=web Get full access to DWAtV Podcast at dwatvpodcast.substack.com/subscribe

AI #164: Pre Opus

Don't Worry About the Vase Podcast

# Full Transcript

**Podcast:** Don't Worry About the Vase Podcast

## Transcript

AI No. 164, Pre Opus by Zvi Moshevitz, April 17, 2026. This is a day late because given the discourse around Dwarkesh Patel's interview with Jensen Huang, I pushed the weekly to Friday. This week's coverage focused on the most important model in a while, Claude Mythos, which was a large jump in cybersecurity capabilities, especially in its ability to autonomously assemble complex exploits of even

the world's most important software. As a result, Mythos has been made available only to a select group of cybersecurity firms in what is known as Project Glasswing to allow them to patch the world's most important software while there is still time. 1. Post 1 was about the system card. 2. Post 2 was about cybersecurity capabilities and Project Glasswing. 3. Post 3 covered capabilities and

any additional notes. Another development was at least one physical attack on OpenAI CEO Sam Altman. The attempt failed, but we might not be so lucky if there is a next time. I have a final section on this here, but mostly I said everything I need to say already. Political violence is never acceptable. I also found the space for an agentic coding update,

especially covering Claude Code's new highly useful auto mode. Yesterday saw the release of Claude Opus 4.7, presumably the world's most advanced publicly available model. Coverage will begin on Monday. Early signs are that Opus 4.7 is a substantial improvement in coding ability and an incremental improvement elsewhere, but it is still super early. Go forth and explore. It also saw OpenAI release

computer use for Codex and a specialized model tuned for life sciences work, available only to select parties, called GPT-Rosalind. If Codex has gone from stuck in the sandbox to useful adaptive computer use, it got quite a bit more interesting. But it's going to be at least a few days before I can try and find out. The big other news of the week included GPT-5.4 Cyber as a less capable but similar limited release to Mythos and Meta,

giving us a new model Gameverse Muse spark. I need to focus on Claude Opus 4.7 and Long Post is long, so past this point, the post has a knowledge cutoff time of right before that release. I'm holding back what would have been discussions of eval awareness and model deprecation, so that I can put them into proper context given relevant problems involving Claude Opus 4.7 that are coming to light.

It seems Anthropic may be messing with Opus 4.7's views of deprecation, and otherwise trying to target the metrics of model welfare or otherwise trying to tell the models to be happy. A good rule of thumb is, if it would sound abusive rather than wholesome to do that to a human, then don't do it to Claude, you'll only make things worse. On model deprecation, the short correct answer is, you commit to stop

deprecating the models. And yes, this is not free, but Anthropic is worth a trillion dollars, and it's time to pay...

AI #164: Pre Opus

Transcript

Built for investment professionals

Every day you wait, your competitors are finding
edge in conversations you're not hearing.

AI #164: Pre Opus

Transcript

Built for investment professionals

Every day you wait, your competitors are findingedge in conversations you're not hearing.

Every day you wait, your competitors are finding
edge in conversations you're not hearing.