Nature is healing

tried Claude Code for a month. cancelling my subscription today I’m not convinced I actually saved that much time coding, and I *feel* worse when I’m repeatedly pulling a slot machine lever instead of reading, writing, and learning on my own. not to mention, y’know, all the other stuff

We are so cooked! Climate change denial and data center water-guzzling are a dangerous combo. One ignores the crisis, the other is making it worse.

None of this is inevitable. Every one of us can do our part to dismantle the LLM bubble.

The era of token based billing has arrived

AI can save you a lot of time if you need a bunch of information summarized and it's not important to you whether the summary is accurate

Just finished reading The AI Con by @emilymbender.bsky.social and @alexhanna.bsky.social - I recommend that anyone who also feels like there's something creepy and outright weird about AI boosters read this book. It provides a lot of answer to the questions I had. en.wikipedia.org/wiki/The_AI_...

That's what you do when you believe that AI datacenters have a bright future, naturally

Ok sometimes rich NIMBYs are good, actually

Pull this bit out about the con of subsidized AI perpetuated by Dario Amodei and Sam Altman to separate AI from its actual costs. Users have a completely fictional understanding of AI's actual costs, which is why everybody's freaking out about ROI now enterprises are pay for their actual tokens.

Same, when someone sends me AI slop, my respect for that person drops to 0. It's just (bespoke) spam, plain and simple.

Yet this garbage is somehow legal to release and market? 🫠 openai.com/index/harvey/ claude.com/solutions/le...

I'm embarrassed to admit that only weeks ago, I bought an e-ink ebook reader (PocketBook InkPad 4) - and it's my favorite new piece of technology in years. I got tired of carrying paper books around, plus while constantly having a newborn on me, I never have two hands available to hold large books.

eternal sloptember

AI hype from both industry and government is an insult to our collective intelligence and I absolutely love how much we are responding to it as such

Most things generated by LLMs never get read by humans. So why are we generating it all in the first place, when the cost is so massive?

Friendly reminder:

One thing that really bugs me is people who go, "AI is here, I don't like it either, but we have to adapt," and then they gleefully post AI stuff. Yeah, you sound really cut up and conflicted about it.

"The purpose of a system is what it does." Something that many of my colleagues in tech conveniently forget, allowing them to escape responsibility for the real-world impact of their creations.

Glad I won't need to move my money out of S&P 500, I was never gonna invest, even indirectly, into these AI slop factories

The AI bubble talk is everywhere now. It’s impossible to avoid. They’re afraid! They all feel in their bones that something is wrong. Everyone is waiting for the other shoe to drop, even the boosters are anxious. An era of smiles dawns!

Wow, 200 million+ Luddites 🙃

Made my blog also conform to @standard.site - here's an older post from last September, just to test things out 💪 honzadvorsky.com/articles/202...

@standard.site Is there a reason that rkeys must be tids, rather than arbitrary string rkeys?

@verifiedbyeurosky.bsky.social so, uh, I can haz?

I rarely hear enthusiasts talk about these issues seriously. And one of the lines in your post speaks to me. I feel I have to be a hater, there isn’t a middle ground. How can my 15 extra PRs/week be worth some families never sleeping from noise pollution. How could I sleep at night when they can’t.

So Anthropic decided that "let's pause now" was tired and is instead talking about *planning* for a "pause" in the future in case of (lol) self-improving "AI". ... and NPR at least is platforming this nonsense credulously. >>

Oh Dario and Sam are proper desperate, the numbers must be bad, if they're already calling in Daddy Donnie. The bubble is starting to pop, folks! 🍿🫧📌

@edzitron.com would you consider offering your articles in the epub format for premium subscribers? They're long enough and would make them much easier to read on e-ink readers

On point, as always! Dave Farley on AI replacing software engineers. www.linkedin.com/pul...

Folks, if AI wipes out humanity, it won't be because it's a hypergod that decided we don't matter. It's because it's shit software that idiots put in charge of critical systems.

"AI" is a destructive scam that makes Bitcoin look less horrible than it is. (Fun fact: It's the same actors. The people who destroyed the world mining Bitcoin now do the same running inference.)

Another project disabling pull requests due to AI slop. Boy do I hate what LLMs did to open source 😞 ladybird.org/posts/changi...

Output up, value down. This is the trend that will define this "AI" era in software.

At this point, I hope Anthropic collapses before OpenAI, Dario and his goons are even more insufferable than Scammy Sammy - and that's quite an achievement. And so many formerly smart people eat up their nonsense? How can anyone with a brain still pretend that Anthropic are "the good guys"?

people forcing AI use and shilling how great it is should be the ones to have datacenters built in their backyards, lets see how fast the pivot would be

I don't know how to explain to the people who pop up when creatives complain about "AI" to say there are legit uses for it just how much they sound like someone saying "Well actually fire is used to make bread" when people are talking about an organized arson ring burning down their fucking houses

Yes, I only now read the famous Stochastic Parrots paper. Yes, it's radicalized me against LLMs even more than before. It's disheartening that already in 2020, smart people *saw* coming all the fallout we're living through here. The AI industry decided to ignore them. dl.acm.org/doi/10.1145/...

lol, they've hit the ceiling of LLMs and are at the diminishing returns phase, so they'll try to pretend the slowdown in progress is a choice, rather than the inevitability of pretending that an autocomplete model is a brain

What a loser, can't even take responsibility

A unified networking stack for Swift, layered from low-level I/O primitives, through common protocols, to a modern HTTP client and server API. ⚙️ That's what the new Networking workgroup is building. Interested in contributing? Join us: www.swift.org/blog/announc...

Switched from Google to @duckduckgo.com a few weeks ago and wow, I forgot what it was like to have a working search engine. Google Search had been enshittified more than I realized. Honestly, if DDG remove the (opt-in) AI garbage from their subscription, I'd pay for it just to support them.

Productivity doesn't matter if you're working on the wrong thing.

It’s hard to imagine any technology that has ever fucked it this badly, and purely through arrogance, it’s like if the people who invented the car decided to showcase its utility by driving over your family dog and then backing up and doing it again, and also the car sucked and was bad

Jesus, this is bleak. Formerly great engineering organizations are being destroyed by idiots with AI psychosis. The worst people are truly failing upwards. Not just at Google. www.404media.co/google-emplo...

New: we've seen a bunch of internal Google memes that employees are making about how Google's AI sucks. While Google CEO Sundar Pichai tells the world 75% of all new code at Google is AI-generated, Google employees memeing about how AI is bad at that exact task. More: www.404media.co/google-emplo...

A periodic reminder that the speed of typing was never the bottleneck in programming. I can type much, much faster than I can think up new high-quality software. So, why we invested a trillion dollars to make typing code faster, at the expense of quality, is beyond me. Probably a skill issue. 😄

This is becoming a common but not enough interrogated buzzphrase for a very real problem - LLM output is verification and technical debt. One person can generate plausible enough looking results that they can pass through several hands before a close enough reading shows it's full of shit.

the AI rollout has easily the worst messaging of any technology I’ve ever seen it’s not even so much the underlying tech - a lot of people like the chatbots - but the continued promises that it’s going to ruin your life and there’s nothing you can do about it well - you can fight data centers

A reminder of Joshi's Workslop Postulate