But ultimately, it's a moot point, because the legal requirement means the books must end up destroyed. Even if the people at Amazon wanted to scan the books in a way that required no destruction at all, it's not currently (legally) possible for them to do so, so they might as well take the easy way out today.

Yes, this is a result of copyright laws. The other commenters are wrong/uninformed.

If it was up to the companies training LLMs, they wouldn't destroy the books: It's a waste of company resources, it's needlessly destructive/evil, it generates bad PR, etc etc. There are essentially zero advantages, other than it is what is required under US copyright law (or at least, it is what their highly paid lawyers believe is required under US copyright law).

Maybe I missed it, but the article doesn't seem to say what the "circumvention"/"loophole" that Amazon used was? It seems like Amazon went through the correct city permitting/legal processes years ago, and now residents want to complain about it years later because data centers are this year's culture war hot topic?

Like if Amazon followed the official process here, what's the problem? At some point you have to declare an approval final and allow a company to move forward with construction.

That said, OpenAI isn't public, so you can't directly short it easily. I guess you could short Microsoft but I don't know if that gives you enough exposure to truly be worth it, even if you do genuinely believe what OOP claims.

Republicans are the only party that in the 21st century acts as though we can simultaneously cut taxes and increase spending and everything will be fine.

As a matter of fact GH even has a red "Less secure" badge on the SMS 2FA in the settings discouraging its use, as well as the following text in the description: "We strongly advise against using SMS because it is susceptible to interception, does not provide resistance against phishing attacks, and deliverability can be unreliable."

A better analogy would be a restaurant deciding not to cater to the 1% of the US population that have celiac disease (cannot eat gluten), or the 2% that have issues with dairy.

Anything Chromium based is tainted. They will not be able to keep out all of Google's shitty decisions because they are not building a browser, they are building a skin on top of somebody else's browser.

If you pass nothing into the API, it doesn't give you an error? Is this even a valid use case - why is it even possible to express this request, should the important inputs not be some kind of required parameter?

If your attempt to use the primary purpose of your API silently broke until your database filled up, that should probably be a big red flag about how likely customers are to make mistakes using your product.

While I mostly agree with your sentiment, I think there is an important difference. Unless you are attempting advantage play (99.99% of gamblers are not, and casinos ban the few that are), there is literally nothing you can do at a casino to make it a positive EV activity. No amount of skill, drive, effort, or anything other than pure luck can consistently generate profit at a casino.

A startup/business, on the other hand, can be effectived by your actions. Luck obviously plays a large factor, but you have some level of control over the outcome.

> Problem 1: It Devours the Context Window

Don't harnesses support progressive discovery these days?

Claude (200K).... GPT-4o..........?

> every MCP server adds a process layer between the LLM and the underlying API

But a CLI doesn't?

------------------

> Measurement: Tool Definition Sizes

> MCP Server: Linear, Notion, Slack, Postgres

Oh, so these are the MCP servers that are examples of context bloat we're going to replace! Later in the article:

> At Quandri we use all three approaches side by side...

> MCP for services without a strong CLI (Slack, Linear, Notion)

I haven't actually checked any sources to confirm there really was downtime on those days, but if we assume those numbers are true 7.8 hours + 1 day is about 1.3 days.

I ask because I'd be pretty disappointed in GrapheneOS over that kind of thing and it'd probably at least partially change my opinion of them, but it's better to validate these types of serious accusations and get the full context.