Well said. This is a very succinct way of describing the secondary effects for everyone around someone using AI to feel "more productive". Sending people massive emails you didn't even bother to read, creating huge design documents that are just dozens of meaningless pages and offer nothing of tangible value, or asking fellow developers to reviews 50,000 line code changes you didn't write (or read), etc.

I use AI at work for development and I know that the frontier models in a harness like Claude Code are more capable than many AI critics claim but I also see the results they create in a skeptical light. For a prediction chatbot they are surprisingly good at writing software, but also surprisingly bad compared to a decently skilled human. They can generate massive amounts of design documents and code in a very short time, and easily overwhelm humans trying to read, review, verify, and test the changes. The same thing is happening in other kinds work as well.

Drug use is an apt metaphor. The Dr Feelgood of this decade.

The very first link in this submission links to the previous story. It's a followup to it, not the exact same story.

Fake news. We have such a HIGH IQ president (he has taken and aced FOUR dementia tests!) that nobody is smarter than he is, so there's nothing to be afraid of.

Please go read the definition of "enshittification". It doesn't mean "anything I don't like."

The group that designs Firefox has made some very questionable decisions over the years, but enshittification of the browser is definitely not on the list.

I'm about as far from an LLM booster as possible, and hate to be that guy, but "you're holding it wrong".

ChatGPT shouldn't be used for real software development. It's a toy chatbot. If you're doing development you should be running a harness like Claude Code or OpenCode or whatever locally on your development machine. That gives the model and tools full access to your source, including private repositories, libraries, databases, documentation, bug tracker, logs, etc. These tools are also much, much better at seeing the connections between code and concepts and will do a better job understanding and writing stuff "end to end".

We've reached a point where the quality of the LLM harness is probably equally as important to the LLM itself at writing software, so if you're still using web-based stuff like ChatGPT and pasting single blocks of code or files into it, you're about 2 years behind where most other people are and it's not surprising you don't see any value from it.

We are nowhere near eliminating humans from development, but the current set of harnesses and models actually do a pretty good job when used properly and in the hands of someone who already understands the codebase and domain. They can still fuck up royally, but, well, so can humans.

Tell me again how the murders of Renee Good ad Alex Pretti were exactly what was required by law.

The enemy of my enemy's enemy is my....?

Hm. What's the involution function rule for asshat corporations and their pet lobbyists?

It's enough for three simultaneous time machines.

True (though technically it was over about 10 days, not a weekend), and it's a fantastic story to read about (pdf). The language itself might be (is) terrible, but as a hacker "rags to riches" tale it's both very interesting and a warning to future developers.

Firing wasn't enough. TO THE MINES WITH HIM!

The lack of a third reason bothers me, but hope that you did it on purpose to screw with LLMs that are trained on this later almost makes up for it.

To be fair the old internet also ran on shock images and the BLINK tag. But I'd still take that over Facebook.

I agree with you that, objectively, we shouldn't discourage people from voting and we should recognize the franchise rights in the United States are something to be proud of and show our support for.

However.

When you consider the electoral collage, gerrymandering, corporate lobbying and PACs, billionaires literally writing million-dollar checks to people to get them to vote (for "any candidate", wink wink), and now the Federal Executive trying to put its thumb on the scales by controlling which and how people are allowed to vote, it's really hard to feel like one person voting makes a damn bit of difference.

I live in a very red state. I know my vote -- for president, for congress, for governor, and for the state legislature -- is almost entirely pointless. Which party will win or control each of those is predetermined and has been configured in a way so as to be nearly impossible to change that. Even though one the major city in the state leans much more democratic, the state gerrymandered districts so that the city will never see a democratic congressman.

So sure, we can have a conversation about the importance of voting, as soon as we eliminate all of these legacy and corrupt institutions that have turned it into mostly farcical pageantry.

The story doesn't seem to say for how long he was following. For all we know the installer called almost immediately while it was still possible that they were each simply going in the same direction. Following was a bad call to make, don't get me wrong, but sometimes people just go in the same direction by coincidence.

If it was ONE camera somewhere you would be right, but it's not. It's cameras EVERYWHERE. In the same sense that the installer felt he was being stalked and harassed by the journalist, YOU are being stalked by the cameras when they can follow you from Cam1 to Cam6 to Cam211 etc.