Failing ever upward

An aside, but the POW MIA flag always looked like the ISIS flag to me

The voice of a convicted child sex offender is important and consequently must be heard

Ghislaine Maxwell denies seeing ‘inappropriate’ conduct by Trump

Wow i guess that clears all that up then

U.S. government takes 10% stake in Intel, as Trump expands control over private sector

I can assume this will go to Green Card holders next, then naturalized citizens.

FBI raids home of John Bolton, former Trump national security advisor

As good a time as any to repost Tony Benn dunking on Bolton

They’ll put her there to make rent control illegal

95% of Companies See ‘Zero Return’ on $30 Billion Generative AI Spend, MIT Report Finds

$30 billion spent proving water is wet, cool

the ai bubble bursting will be a small mercy, but a welcome one

Flower Hill operates as a democratic, diverse, communitarian, and heavily militarized federation of villages attempting to live and prosper in harmony in spite of the imperialistic aggression from their neighboring territory and archenemy, the Weasel Empire.

It’s giving

Janet Yellen - tiny little economic hummingbird, but STRONG, flaps her wings and the rates just DROP - faster than light, faster than gravity, faster than Einstein (never ran the Fed!). Powell? TOTAL disaster. Probably working for China, maybe Mars. Janet made money cheaper than a BigMac in 72, and I’m thinking about bringing her BACK, maybe put her in chargeof EVERYTHING

Come on, Bryce. There are a lot more important problems than Sri Lanka Palestine to worry about.

Thanks comrade, it’s the utter worst lol…not only is the job market terrible, I’m at the point in my career where working for a company that is even remotely reasonable in what they do and how they do it is the bare minimum qualifying factor. Turns out….there’s not many of them…

Sadly Yegge now works for Sourcegraph, and as such, is one of the most hyperbolic LLM shills out there. Here’s the article:

How do you know if you’re doing AI right at your company? We’ve noticed that the companies that are winning with AI – the ones happy with their progress – tend to be the ones that encourage token burn. Token spend per developer per unit time is the new health metric that best represents how well your company is doing with AI: an idea proposed by Dr. Matt Beane and playing out in the field as we speak. I see companies saying, “If our devs are spending $100-$300 a day, that’s much less than paying for another human engineer. So if AI makes our devs twice as productive, or in some cases only 50% more, we’re winning.”

There’s an implicit assumption here that tokens == productivity. We don’t have any data to back this up yet, so if you’re about to have a seizure from wanting to argue with me, this would be a good spot to pounce and call bullshit on every word of this post.

I was working on a take home project for a job I applied for, and I burned through my paid for Claude limits for the day soooo quickly - I wasn’t even using Claude Code or Cursor. Bear in mind this is a tiny, toy project, too. I remember the Steve Yegge article where he talked about companies incentivizing token spend and using it as a metric of assessing the performance of a developer (yes, tech execs are that detached from reality). I think wheels will fall off that wagon very quickly, right at the same time as the gelatinous soup of unmaintainable GenAI code begins to become a significant problem for companies, too.

Lmao, the free money is drying up for AI companies and they’re enshittifying hard. From the massive rate limiting on Claude’s APIs, to this - the free ride is over, and the largest capital expenditure in the history of tech is producing fuck all to show for it.

As ever, Ed Zitron has been great on this:

In the event that they are representative of the greater picture of Anthropic’s customer base, this company is wilfully burning 200% to 3000% of each Pro or Max customer that interacts with Claude Code, and in each price point’s case I have found repeated evidence that customers are allowed to burn their entire monthly payment in compute within, at best, eight days, with some cases involving customers on a $200-a-month subscription burning as much as $10,000 worth of compute.

…

I am, however, unsure what restrictions Anthropic could make that would ease the burden on its infrastructure without also hemorrhaging users. I am deadly serious when I say that almost any interaction with Claude Code loses Anthropic money, and anyone using it for even a few hours a day is guaranteed to burn through the majority of their subscription revenue in a couple of days.

This is not a sustainable business model, but also suggests that the violent and sudden changes to products like Lovable, Replit and Cursor are indicative of the real economics of these models, and the likely future form of products offered by OpenAI and Anthropic themselves — opaque rate limits, random service outages, constant demands to upgrade and unhappy customers churning every hour.

Edit: just played with it, the AGI is here folks