toebert

Sure, but I’ve only started recently so not sure how it’ll hold up. I’ve managed to get my bank transactions in there and my investments as well, and for a brief period my crypto too. It all needs external integrations though (or you can do it by hand).

It does have tags and rules to automate some categorisation. There are a few people maintaining it but it obviously came with some tech debt too (it was originally developed by a company and called maybe, they open sourced it when they gave up on it).

China is the same, but the western world is in a race to the bottom so at least we can wave to them as we pass them on the way down.

I don’t know the context around you getting banned, unless there’s some specific rules you violated. I am not in support of that, but it’s also not the focus of my message.

I disagree with development having to go that way. If anything, the hatred towards ai is a sign that it’s actively not sought after, or at least not with LLMs. If they managed to develop actual AI that is on par with senior engineers, maybe? But we don’t have that. What we have is faulty and inherently flawed. Why would we have to push ahead forcefully with it…?

I didn’t include a list of why ai is harmful as the post was already long, but displacing workers is just 1 point.

• massive waste of resources (as in water, electricity) for tasks which can already be achieved without AI for a fraction of the compute cost (think, search engines as an example). Also consider the environmental impact here in a society where a lot of our power still comes from burning fossil fuels.
• a war on consume hardware (all compute components “sold out” for 1-2 years ahead making everything expensive for average people)
• destruction of the workforce pipeline (even if only junior roles got displaced by ai, we will simply not have a pipeline of new staff to step in once seniors had enough, in any industry this is catastrophic, especially when the machine doing this is not actually able to fully replace staff)
• building a dependence of closed source subscription based tooling or end up locked out of your own codebase because it’s infeasible to do it without once you started
• theft of intellectual property ignoring all licensing for training data, or companies selling individual contributions
• the entire thing being funded by imaginary money propped up by a circle of loans driving us towards yet another financial collapse across the modern world

I’m sure there are even more.

Not all of these are the fault of the technology, but I’m more than happy to throw the entire technology and everything around it under the bus if it means it makes it easier for people to unite against these companies - which I think it does.

Saying “it’s a tool and provides value” is like saying “force feeding chickens in a tiny cage” is a tool that provides value. True? Yes. Valid? No.

Even if that were true (and in some rare cases it probably is) the machine is trained on stolen data, ignoring all licensing or companies selling people’s contributions without their approval - and that’s just the tip of the iceberg.

To call it slop is a great way to discredit it and to not support an unethical business/technology.

I don’t think everything is getting called ai slop, but I would say if any part of your project is ai slop (like your “lazy uis”) I’d also immediately lose trust in the entirety of the project, especially if it’s intended to be around security. I do think most projects that use AI for code generation are slop though, I’ve seen far fewer examples of good use (i.e. where the output looks human written because the operator reviewed and refactored every part of it, or where it was used to write small parts of functions rather than entire functionalities)

Your last sentence I think provides a great argument for why people here (and more and more broadly in engineering) hate on ai generated code in general. It produces such vast quantities of code (and often unnecessarily) that it becomes infeasible for a human to review it, immediately requiring us to place trust in the machine to both generate it and review it, and to continue maintaining it while the human operator probably does not even have full understanding of what’s changing. A machine, that we all know hallucinates and generates often low quality garbage, including severe security vulnerabilities by design. According to GitHub, your project has millions of lines of changes on a weekly basis in the earlier days, that does scream slop to me.

Last, AI is more and more hated due to the increasing number of horrible impacts it has on our world, personally I’d not support AI generated projects just on that principle alone.

They’re harder to eat

800 pounds for 64gb ddr5, what a time to be alive. Could get something a bit cheaper but it’s still crazy prices.

Even with cash we’re at the mercy of a country, if they fuck up their economy and hyperinflate it, money is gone anyway.

The only way forward is to carry around stuff that has intrinsic universal value. The currency of the future is potatoes, start stocking up.

I never understood the argument of “if it’s not in the UI you may need to use a command to achieve it and it’s scary”.

On windows, if it’s not in the UI you have to use either a powershell command or update the registry to change it - which are both a very similar experience.

The only difference I actually see in this point is that Linux has a lot more options.