Home Creators Posts Import Register Favorites Logout

Downloads

Content

Featuring the tale of a liar trying to sell something to professional liar-wrangler Michael Marshall! And some other tales.This is the audio only version. Video version coming soon.YAHTZEE!

On this week's episode: We’ve always been at war with Eastasia ... An old Who’s Woo subject gets in touch with an opportunity too good to turn down ... And presidential hopeful Tucker Carlson wants a house to go with his replacement theory.

Watch the patron-only ad-free video version at our Patreon site or use the private link here: https://youtu.be/N7jou1ePgS4

Headline Sources:

Trump basically resorts to threatening Iran with imaginary missiles: https://abcnews.com/Politics/crucial-us-missile-stockpiles-extremely-low-official/story?id=135428997 https://www.washingtonpost.com/national-security/2026/08/05/trump-hegseth-clashed-camp-david-over-iran-missile-depletion-concerns/

Playbooks, plagiarism and a pig’s head: new claims surrounding a star Cambridge professor: https://www.theguardian.com/education/ng-interactive/2026/aug/01/playbooks-plagiarism-pigs-head-new-claims-surrounding-cambridge-professor-jason-arday The takedown of Jason Arday has overjoyed the right, and must be a wake-up call for the left: https://www.theguardian.com/commentisfree/2026/aug/06/jason-arday-dei-race-issues-university-of-cambridge

Comey prosecution is absurd and Trump reminds us that he was a Russian asset: https://www.nytimes.com/2026/07/29/us/james-comey-seashells-trump-secret-service.htmlhttps://www.axios.com/2026/08/05/trump-russian-asset-comey-fbi-investigation

Fauci takes the 5th: https://www.nytimes.com/2026/08/06/us/politics/fauci-contempt-of-congress.html

Michael Marshall has a zoom call with an unsuspecting Brian Rose: https://www.facebook.com/reel/1580651800303373  

TUCKER CARLSON puts out 10-point manifesto in possible bid for president: https://www.axios.com/2026/08/06/tucker-carlson-third-party-israel-trump

Comments

Max

How do we watch the call between Marsh and Brian Rose?

Talyn

Um guys, can we have a whisper fight real quick? Quit telling this administration they're running out of munitions. They might nuke Iran if they figure out that's the only way to win. You really think they wont?

zedd_d1abl0

There is limited cause for concern about the HuggingFace hack. It was a novel attack, but the escape from the sandbox isn't that scary, depending on the sandbox they built. Also, your bigger worry should 1000% be the level of competence it takes to be an administrator of a system. Being an IT admin in most places is just accepting the pay check There's no real formalised level of acceptance or criteria for a Sysadmin.. I've personally dealt with Sysadmins that didn't know what firewalls were, how DNS worked, or what ping does. The best way I've heard AI described is that it's "Better than average". If you're an average employee, you've got a rough future ahead. If you're even slightly better than average, you're good. Sure, AI can do things quickly, but by its nature, it'll repeatedly fuck up the same thing because of how it generates its values.

Alex

I keep wondering if this administration can make this war even stupider but they always rise to the challenge

Tinkerton

Casually dropping in an Oxford comma joke has made my week, possibly my whole year 👏🏻🤩

Dahak

Ahh, sad I can't watch the Marsh call since I'm not on Facebook (and will never be damnit!)

Ricky Burnett

Great show. I also wanted to comment that you shouldn't be that concerned about the HuggingFace attack and the subsequent Anthropic announcement. It was an interesting attack, but those happen all the times. I'll try to set the context to help you assess these announcements: There are currently 10's to 100's of thousands of human security and attack experts who are experimenting with the tools to hack into systems every day. They currently have little to no oversight given that most of them are literal criminals. These people are doing more than just experimenting. To bring AI into it. AI is probably faster than humans are finding new ways to use these tools. That's what is meant by a novel attack. There are a couple hundred of these per day (google CVE). But the tools are the same and are well known. Grossly simplifying, the victim can just start turning things off when the attack is spotted. If AI was 1000x times smarter, they would hit the same limits as today. In my opinion, the canary you should be looking for is scale. If one of the big AI companies starts deploying a huge number of agents, and these are breaking out of their sandboxes 100s of times a day, there is a risk that so many of them will be out there and they will be moving so fast that admins can't respond. At that point there is a real risk of damage before agents get caught and victims start shutting things down.

The Skepticrat

Heath here. I put the link in the notes above. The video is priceless!

LisaLisa

I wore a tshirt that said "86 45" during the first administration. Secret service must have been very bored following me around waiting for me to make a move