>> Look, you've drank the Kool Aid

You told Claude to "do a report with some complex math to analyze user behavior", lets see your prompts. And then you had to show "the output to the data scientist" to determine if it was correct, you couldn't figure it out on your own. I'm thinking you don't have the skill or the patience to use these tools.

AI isn't a magic wand. You can't just wave your hands in the general direction of what you think you might want. You have to be specific and break the task down into manageable pieces, just like you would with a human.

The various AI models have implemented about 200k lines of excellent, functioning code for me in a variety of projects since last spring so if this is Kool Aid I'm not complaining. And I'm not alone, there is very broad adoption as the linked article confirms and there is plenty of other evidence. The value of software company stocks recently fell off significantly on signs that AI can replicate much of what they do. Same with legal;
https://ancillary-proxy.atarimworker.io?url=https%3A%2F%2Flegaltechnology.com%2F20...

I have to think that most people posting here don't use Claude for coding, or at least not recently. It is amazingly effective, and has steadily and noticeably improved month over month. From the article (which many apparently didn't bother to read);

"Provide Claude (the CLI) an input such as a spreadsheet, a codebase, a link to a webpage and then ask it to achieve an objective. It then makes a plan, verifies details, and then executes it."

And that's what it does. It can describe in detail what it would accomplish and write it out in a document for you to review. You can make revisions to the document directly or just tell it what to change. You can tell it to implement things in phases so you can check for correctness step by step. The code quality is usually quite good, and if there are bugs you can just paste in the error messages and it will fix them.

This not a bad thing, it is empowering. An average programmer turns into a 10x. An excellent programmer becomes a 100x. What's not to like?

I get the strong impression that people posting here do not use these tools.

I routinely code with AI assistance, and I can choose from 2-3 dozen models as I go. The cost per prompt ranges between free to a nickel or two. In general you can get a lot of work done for you for under 50 cents. People working in corporate environments get team-level access for a few $hundred/month and can burn through all the prompts they want.

Claude Opus 4.5 has been the premier coding tool for the past several months by far. The prompt cost has reflected that, and in my experience it has been well worth it. So when I see a 4.6 come out its great, but I want to see a reduction in the usage price of 4.5 which is wickedly good.

"sensitive information—notably command data for European satellites—is unencrypted, because many were launched years ago without advanced onboard computers or encryption capabilities."

Pretty surprising that the command data isn't encrypted, that's a severe vulnerability. I don't know how old those satellites are but it wouldn't take a very "advanced onboard computer" to implement some basic encryption.

"risks including competition from AI-native firms and clients building their own solutions in-house"

Apparently investors think the software companies could be overtaken permanently.

>> If you have an idea, it can be done with vibe coding today if you have the intelligence and creativity to do it. Simple as. If you don't, you can't - and won't.

I think there are a couple of definitions of 'vibe coding'. One is where a person who is not very familiar with writing software gets to have an app of some kind created by just going through an exploring sequence of pretty vague prompts. It's very cool that people can do that now.

Another definition is where an experienced programmer can get a good LLM to do all the grunt work that would be involved in anything he can dream up, or "vibe".

It used to be that you had to marshal your time. Stuff that might be extremely useful or interesting to do might take too much time and effort to justify. Not any longer. You can condense a month of work into a couple of days. If things didn't work out it isn't a big loss. It give tremendous freedom to experiment.

That's where the innovation comes from.

>> outside, enjoying the outdoors, moving around

You might like it for a few days as a nice change away from the monitor/keyboard, but it is mindless work in good weather and bad. One guy's job was to move a handheld device around on the ground to pinpoint the location of the drill head. Do you really want to devote a major portion of your waking hours to that every day, month after month? And then there are the guys who dig holes out by the curb with shovels (or pickaxe as required) for the junction boxes.

AT&T ran cable all through my neighborhood a couple of years ago. I got a good look at the kind of very unpleasant work it is. They use dirty, noisy tunneling machines to burrow under roads, driveways, and sidewalks. They are standing out in the weather (in this case more than 100F in the shade). It was all done by contractors. Probably there is some optical fiber somewhere that needs to be spliced, but in this case they were just pulling what looked like ordinary copper cable.

I'm very skeptical that they were getting paid $35/hr.

The article gives some details.
“We had to totally rethink the computer program instructions,” says Blaauw, “condensing what conventionally would require many instructions for propulsion control into a single, special instruction to shrink the program’s length to fit in the robot’s tiny memory space.”

So I'm suspecting that it has just a few high level instructions. Apparently they can issue commands to it with modulated light, but it can't respond that way at present. Probably they will build on this platform and add capabilities.

“To report out their temperature measurements, we designed a special computer instruction that encodes a value, such as the measured temperature, in the wiggles of a little dance the robot performs,” says Blaauw. “We then look at this dance through a microscope with a camera and decode from the wiggles what the robots are saying to us. It’s very similar to how honey bees communicate with each other.”

But it seems like steering these will be a challenge. The video in their paper shows a prototype unit successfully propelling itself through a narrow channel. When it leaves the channel it turns sharply and flips over.

Utter bullshit.

If you are an IT worker most of your work hours are going to be spent sitting in a chair in front of a keyboard and a monitor. Maybe just a laptop, maybe several screens, and you gotta come up with some software there. If you have to drive in and work in a bullpen office environment it will be much the same scenario. People are mainly just trying to get their work done with the computer interfaces in front of them.

But on the other hand there reportedly are those who work multiple jobs remotely without the employers being aware of it. And these days maybe it is easier to do. If you know how to use AI assistance effectively you can generate the coding output of 2-3 unboosted humans without undue effort.

I like your style. But how's the missus?

>> Chinese efforts are absolutely on par with what is going on in the US

I use these products daily for coding and in my experience the Chinese LLM's are significantly inferior to Anthropic, Google, OpenAI, etc.

>> The electricity that lights our homes. The fuel that moves our cars

Relatively little oil is used for producing electricity, roughly 2-3% of global electricity generation. Electric vehicles accounted for about 25% of global new car sales in 2025.