My first Substack post (of many) has finally dropped! Come and get some. I laid out my official 1 year prediction for AI. TLDR: the bubble pops. benjaminhorne.substack.com/p…

He just broke the first rule of AI hype posting. He’ll be hearing a knock at his door any minute now.

People like this think they stand on a moral high ground because they proudly cut people off when they make mistakes (even very ugly ones, yes) but this is very transparently a panicked act of tribal self-preservation, and has nothing to do with right and wrong. Her sense of morality is purely performative, hollow, and like many chronically online women such as herself, based completely off of probing for whatever she thinks is the latest aesthetic consensus of the hive mind. She is not a real person. She has no thoughts of her own. There is no real ground beneath her feet. She would become pro-child murder tomorrow if the comment section suddenly moved to that position.

“If I don’t build it, China will,” I said to my wife, tears streaming down her face, as I walked out the door to work on my backyard meth lab for the 5th weekend in a row.

Confirmed.

Semi is shitposting, and you’re bearish?

Meanwhile, this is what Chinese AI researchers look like. We are cooked.

I sleep better than all of you because I don’t manically use technology to track it.

There is an argument to be made that this could turn out bullish for Anthropic and the AI industry if the US Gov realizes that, if they’re going to crush companies’ business/revenue by blocking them from cashing in on selling something they spent *billions* to make (Fable), it’s only right to find some way to make up for it somehow on the backend. This is especially true for the obvious strategic/national security implications as well. However I’m not getting the vibe that this is going to happen, it feels more like a “fuck you” from Hegseth et al.

I suspect at this moment there is a super smart and talented young person out there somewhere building software that is going to do to Adobe Premiere what @zoink did to Adobe XD.

Yep. GLM 5.2 just dropped and apparently it's on par with Opus 4.8. It is extremely hard to imagine a scenario where US gov doesn't reverse the recent Anthropic/Fable ban, unless they are completely insane.

I wonder if he's met up with @ChineseStalwart yet..

The honeymoon is over indeed. Big time ARR collapse incoming.

I'm not actually an Elon hater per se. He is no doubt the greatest entrepreneur of his generation, and probably greatest all time. Credit where it's due. I also appreciate his perspectives on many things, and his relentless & unapologetic optimism about technological progress. That said, he is an absolute fucking retard in many ways, too. He lies quite a bit (even about video games, for fuck's sake), and his knowledge of political matters is that of a 15 year old 4channer. And per my tweet below, 90% of his fans are cringe and annoying as fuck, and weirdly obsessed with him despite the fact if there ever did come a time to escape to Mars, they ain't getting a seat on the spaceship.

Correct. But... this guarantees that the AI bubble bursts.