I think the problem is the techno fascist oligarchs that are peddling the snake oil that LLMs will wipe out all white collar jobs tomorrow. Managers usually answer to C suite, and the C suite is salivating at the idea of laying off 80% of staff

Not sure homicide fits into “small crime” category.

No, but it serves as a sense check on the other data. If the official stats were bogus and crime were spiraling out of control in London, it would be somewhat surprising to see homicides going down. The fact that one of the most objectively measurable crimes is going down lends some additional credibility to the statistics indicating that this trend is also being seen across other crime categories.

People are outsourcing even basic sql to AI? Do you do any thinking for yourself still? I’m super curious what the AI brain rot will look like even 5 years from now when some senior dev can’t do a join without AI

Well, we used ORMs for a decade and treated our code as schema.

Now we hooked all that up to agents.

I can still do a join and often write sql queries by hand to execute in a shell.

> Now we hooked all that up to agents.

I and many others are not in your we

Or quality

Actually synthetic training dats is better, thats why the new models are all better at design.

If synthetic data is so much better then what are AI crawlers still DDOSing everyone for? Are they stupid?

Mostly. I had the "AI bot tsunami" problem on my own personal site and blocked a bunch of bot user agents in robots.txt. Most of them were from companies I had never heard of before. The only big AI name I recognized was GPTBot from OpenAI.

It’s incredibly tiring to see this narrative peddled every damn day. I use opus 4.5 every day. It’s not much different than any previous models, still does dumb things all the time.

Same experience - I've had it fail at the same reasonably simple tasks I had opus 4 and sonnet 4.5 and sonnet 4 fail at when they aren't carefully guided and their work check and fixed...

What is bad exactly about the trajectory? They are constantly adding features and none of feels like bloat, it just happens to be a much younger editor so iteration is rapid. I welcome it compared to the monstrosity that Jetbrains has become.

Yeah, it just reminds me of the early days of VS Code, where features were constantly added and it was fun at first, but they didn't stop and eventually it did feel more sluggish and bloated. Sometimes I'd have to spend time fixing or re-configuring something just because I opened the editor and the daily auto-update did something annoying. It might not happen with Zed, but it seems like a very similar approach to development.

I find the mental gymnastics like this around AI discourse really confusing. Taking a still video of the ocean isn't "creativity". And if a video of an ocean is truly creative it's likely not attributed to "nature".

Taking still photos of anything is widely considered a creative activity. There's usually only a small amount of creativity involved. Sometimes more. Same goes for AI generated anything. You can have low effort photos and high effort photos, low effort AI generated pictures and high-effort AI generated pictures.

AI just lets you get a better result for less effort (for some definitions of "better"), just as a camera lets you get a better result for less effort than a paintbrush and canvas does.

A camera doesn't produce a painting though. Your comment doesn't make any sense.

It produces an image, just like painting does. It does so quicker and "better" than humans. But it also requires less creative input and allows less creative freedom.

> a camera lets you get a better result for less effort than a paintbrush and canvas does

Only if your goal is photorealism, which is one of the least interesting forms of art.

How do you determine that a typical job is busy work? While there are certainly jobs like that, I don’t really see them being more than a fraction of the total white collar labour force.

Yeah that kind of thinking is known as “doorman fallacy”. Essentially the job whose full value is not immediately obvious to ignorant observer = “useless busy work”.

I genuinely despise READMEs that are just AI slopped emoji hell. Can we not put some thought and effort into things anymore...?

As a twitch reaction I’m inclined to agree, but I’ve found the quality of AI generated readmes typically far exceed those created by humans. Yes, some emojis and fluff are to be expected, but also great structure, fully documented CLI arguments, loads of examples and easy to get started guidelines.

It does however indicate that likely the code is also fully or at best partially written by the same LLM.

The entire project is just a paralell wrapper for a python library

Well by not opening the blog post or whatever page that nicely explains the JavaScript sort with examples, you just deprived them of page views and probably income. So what will happen in 5 years when you’re searching for human written and thoughtful content on something more complex and all you get is slop?

You haven't really been getting 'human written and thoughtful content' for a vast swath of search topics for probably 15-20 years now. You get SEO-hyper-optimized (probably LLM-generated for anything in the last 3 years) blog spam. In terms of searching for information and getting that information, there are a lot of topics where an LLM-generated result is vastly better just by virtue of not being buried inside blog spam. The slop ship sailed years ago.