selfhoster11

Please add Internet Archive's bot to your auto-allows, at least. Their bot is presumably well behaved, and for public benefit.

No, the problem absolutely is identity verification. No KYC should be needed for any kind of online API service, period.

It doesn't take much hardware. I have run larger models.

I have seen such drives. ATAboy promises to be a nice option if manufacturing and populating a PCB is not too expensive.

It's far better than everyone has nukes than just a few people who are highly interested in ruining your mind and/or finances. Governments, crime syndicates can pay for HHH-less AI.

Counterpoint: B corporations.

It's clearly possible for companies to self-impose safeguards: ESG/DEI, Bcorp, choosing to open source, and so on. If investors squeal, find better investors or tell them to put up with it. You can make plenty of profit without making all the profit that can be made.

I don't think the parent comment was saying it can be solved, only that the LLM paradigm is better at dealing with complexity. I agree that they are not great at it yet, but I've seen vast improvements in the past 3 months alone.

GeoIP this nonsense. Legal liability is solved as a "good-faith effort" and those living in jurisdictions where this doesn't apply (or use a VPN) don't need to be stripped of privacy.

Murder is immoral. That much is certain. ICE murders people.

FWIW, you aren't alone man. Stay strong.

Except the parent comment said you can stream the weights from an SSD. The full weights, uncompressed. It takes a little longer (a lot longer), but the model at least works without lossy pre-processing.

Try doing this as a normies without technical guidance. Technically correct, this time, is not the benchmark.

OpenAI does KYC. I refuse to deal with that.

I disagree. Music needs to be archived just as much as books and other publications. Yes, there is heat, but lots of people support them.

Good thing that there are some people to counterbalance that, then.

Not the point. The point is that nobody who doesn't know him knows whether he can be trusted, and the ISO itself might be prepared on a compromised machine unintentionally. This is very much a supply chain issue.

The users.

They already are paying off. The nature of LLMs means that they will require expensive, fast hardware that's a large capex.

So only those with money should be allowed to play with tools? If you have no income, are unbanked, or are in an area with low earnings, that compiler is out of reach. And then you're back to piracy.

You can do that yourself with opencode and a $10-15 subscription to a good model.

I used to be a skeptic, but this month the pairing of tooling and models I discovered is finally good enough to build a static Go app with a TypeScript front-end, along with all the deployment and compile steps, despite knowing nothing about either of these.

Insufficient or bad wrapper will hobble a model. An insufficient model will mean the wrapper cannot help much. Both have to be in place, and this month they finally came together for me to create ready-to-go stuff.

Heads up that this is "more true" for non-reasoning LLMs. Reasoning gives an LLM a lot more runway to respond to out of distribution inputs by devoting more compute on understanding the code it's changing, and to play with ideas on how to change it before it commits.

1. You can run decent local AI now - see /r/LocalLlama. You pay the electricity cost and hardware capex (which isn't that expensive for smaller models).

2. Chinese APIs like Moonshot and DeepSeek have extremely cheap pricing, with optional subscriptions that will grant you a fixed number of requests of any context size for under $10 a month. Claude Code is the bourgeois option, GLM-4.7 does quite well on vibe coding and is extremely cheap.

Gemma 3 27B, some smaller models in the 8-16B size range, and up to 32B can be run on hardware that fits in the "consumer" bracket. RAM is more expensive now, but most people can afford a machine with 32GB and maybe a small graphics card.

Small models don't have as much world knowledge as very large models (proprietary or open source ones), but it's not always needed. They still can do a lot of stuff. OCR and image captioning, tagging, following well-defined instructions, general chat, some coding, are all things local models do pretty well.

Edit: fixed unnecessarily abrasive wording

Far less than you'd think for local LLMs.

Kimi K2 Thinking. Tell it in the system prompt to push back.

It's not that hard. Stationery shops do exist, even online if you don't have one locally.

Better this than 300th React.JS bloatware of the year.

You may want to use the new "derestricted" variants of gpt-oss. While the ostensible goal of these variants is to de-censor them, it ends up removing the models' obsession with policy and wasting thinking tokens that could be used towards actually reasoning through a problem.

I mean, yes. Very much so. People should be upset about a relatively affordable hobby getting to this point.

In the US, yes.