devin

I don't think pairing, nor TDD come anywhere close to the abdication of responsibility that agentic programming does.

The good devs I know avoid dependencies like the plague.

I had a similar take, but I was wrong.

It's going to be funny/sad to watch people who have deeply misapplied LLM tech discover "oh gee, this is pretty expensive. I guess I can just use ML?"

Their lot (VP-level certainly) rotates in and out, often with 3-4 layers of middle management between them and the crews executing, which makes direct assignment of blame exceedingly difficult. Every 3-4 years you get a near-complete rotation of leadership in an area of †he business, and any memory of what anyone was trying to achieve before is forgotten, either intentionally or unintentionally, and then you do it all over again.

It's different to me from text written by another person, because it is awful in a way that even a bad human writer can't replicate without first performing a lobotomy on themself.

Did you forget to put an /s on this?

> The architectural differences that made a console more compelling than a standalone PC are largely gone.

No, the main driver is still there. People do not want to build and maintain PCs. Your 14-year-old nephew doesn't want to do it, either. People want to turn on a box and play a game. I am technical and I want absolutely nothing to do with maintaining a PC. I'll take the console experience every single time.

Sure, but it's also feels like incredibly short-term thinking. Embarrassingly so.

Then after your hour long drive it happens? Great!

Not sure what your point is.

You can just restate the question then to say: If the models are so good, why can't they run a business?

I hear you, but also I'm not so sure. I remember Log4j, heartbleed, etc. and they were just as serious, but I don't recall people just going dark immediately. Scripts to validate whether or not you had exposure were built, people ran them, and if you did have exposure you started checking to see whether or not it was exploited. This timeline for this seems wildly different to me.

I think the article is suggesting that a sober risk assessment was not done by most folks period, not that they merely perceived the risk differently.

Same. People are drunk on "agentic" and reality is coming for them. It's also coming for the holdouts who say it has no effect.

Like a lot of tech before it I think the overriding theme is misapplication during a hype cycle.

Fair enough. I call it out because I know that some docs are trying it for people who have a history of substance abuse, and as a first line instead of jumping straight to benzos because of their risk profile. A lot of the reports I've read about hydroxizine are "it just knocks me out, which is I guess better than panicking", but also "it just knocks me out" seems preferable to risking dependency and the hell that benzo withdrawal can become.

Edit: Granted, falling asleep is not really conducive to being able to work. So I imagine there a number of people where "it made me fall asleep, but I can't do my job like that" is a non-starter, and short-term benzos is the only way forward.

Isn't it fairly standard these days for docs to offer hydroxizine instead of a benzo for panic attacks?

It blew my mind to enter college and peer review other people's papers. Up until that point I understood that there was a range of capability with writing, but I figured that in college courses the floor would be pushed up. I was dead wrong, and it was shocking to me that these folks had managed to get all the way through high school writing like 8th graders.

ETA: This is over 20 years ago, btw. I hear it's gotten materially worse. That terrifies me.

And Clojure. Perhaps these aren't considered low resource. I had good results in Clojure over a year ago.

I spent years programming for myself, for free. I did this because I loved it, not because I was interested in the VC game.

In some ways it feels like exactly what you'd do if you wanted to tank your product over the coming years. I look back at the last 20 years or so, and what I see over and over is that one player in a space gets huge, keeps adding features, loses sight of what makes them a good market fit, and then 10 small upstarts swoop in to bring back the opinionated or simple view again. One or two win, and then eventually they repeat the same mistake.

I'm not trying to buy a labyrinth.

I’m aware.

The fact that multiple replies to my post know this fact only serves to prove my point though. :)

I'm a lisp nerd, but why do you say sexps are the answer here? I think the point of a lot of configuration formats is that they don't want their configuration to also be a full-fledged programming language.

1. Nerds like Japan.

2. People with money like acquiring passports.

3. This community has above average representation of both.

How much later? Cryptographic systems certainly have methods for allowing revoting up to a deadline, but endlessly mutable voting is insane. It's also not really any different from today. I can say I didn't vote for candidate A 12 months after the upcoming election and it means... well... nothing.

There are a lot of folks who, whenever this topic comes up, seem to have no knowledge of the systems which already exist and have actually been used in real elections already. For instance, ElectionGuard has been used in several real elections in the US since 2020[1].

[1]: https://electionguard.vote/

The reply with "many of us..." is so dismissive that it's hard to believe you understand what I'm saying. You act as if I haven't been using these models on the daily for real work since the beginning. Note the word "concise". Note the rest of what I wrote after it, namely the part about how there is often a bunch of accretion of prompts that leads to any output resembling the one I posted.

You seem more concerned with shining your AI badge than understanding what I wrote.

I don’t think you followed the point being made.