I’m sure I would recoil in terror at the code I wrote but I owe a lot to VB6.

Very common phenomena that is discussed frequently in the souls community.

No, AI code review doesn’t help. Claude can’t even give me correct line numbers 80% of the time, literally just makes them up, and more than half of it is false positive BS anyway.

Should we mandate IUDs for all women and require abortions for “accidents”?

Mandatory vasectomies for every man and only reverse it when they are rich enough? Or perhaps we chemically castrate them once they hit puberty?

What’s your preference?

One of these days Anthropic is going to be compromised and we’re all gonna be f*cked.

Do these refusals still happen if you’re using an API key instead?

So I suppose Anthropic lied to him?

Mitchell said on X if he was in charge, he would shut down copilot completely. It might bring in revenue but without core GitHub, Copilot also goes to zero and he’s totally right about this.

It makes absolutely no sense for Microslop to have Copilot when they have such a huge stake in OpenAI and they have Codex. The price hikes for copilot tokens and the massive backlash proves this is not a good business model. Just shut it down and get back to fundamentals.

AI is a cancer and it slowly hollows out everything it touches. Corporate greed is merely an accelerant.

You might say “it’s merely a tool”, but it’s impossible for me to divorce the realities of the terrible harm these generative AI tools are doing to the world and it’s immensely selfish to be like “well it generates slop code for me good enough so I don’t care about all the other bad stuff”. So I just suffer in silence at work because I’m forced to use AI to generate code even though I probably spend more time just reprompting then I would just writing things by hand.

The worst thing is needing to babysit third world contractors they gave a Claude subscription to - the worst slop I have ever seen. Give me a meatsack junior right out of college with no AI over this garbage any day of the week.

And if that wasn’t bad enough - we had to fire our lead CSM the other day because he secretly reverse engineered our private API with Claude and built something that he put unauthenticated on the public internet that leaked a bunch of customer data. Absolute nightmare dealing with the fallout of this that didn’t need to happen. And if that wasn’t bad enough he tried to extort money from us because the thing he built was supposed to replace one of our vendors. This was by-far the worst case of AI psychosis induced Dunning-Krueger I have ever seen. He’s actually going to try to sue the company now for wrongful termination! I bet Claude told him he was absolutey right!

It’s not like I can just find another job either, the market is sh*t and everything else is forcing these tools anyway. So yeah, thank you SV for destroying the tech labour market, suppressing wages, and making the already deteriorating state of software quality even worse.

Frankly, the ROI of these tools is murky at best and it’s gonna get worse as they raise prices. I don’t see what benefit they actually bring beyond a hit of dopamine when it near-shots a task. I have yet to see any killer app that feels truely like AI is doing something worthwhile. And all the startups I see that are AI first are just selling AI orchestration to other AI startups. BORING. They tried making an AI MySpace and it was a disaster, so what CAN these things actually do?

OpenClaw is supposed to be some revolutionary thing but nobody actually does anything truely novel with it. Oh it aggregates your news for you? Congratulations I have an app that already does that without AI and it doesn’t cost me per token.