If your tiger had full remote supervision capability and a remote kill switch, its ability to go rogue seems like a choice you’re making, not an accident.

Why do you ask questions with obvious ulterior motives though?

Not for the purposes of safety and sanitation. Dilute even 15% ABV wine (the high end) with contaminated water and you end up with a contaminated low ABV wine-flavored drink. Natural fermentation doesn't generate insta-kill levels of alcohol, it was the boiling of the wort or must that made dodgy water safe, the fermentation just kept it that way for longer.

You're clearly out of your depth on this topic.

Perhaps this is placing a bet that Kessler syndrome is a real threat to surveillance capabilities and they want these in case the satellites go down.

Extract a sugary liquid from fruit or malted grain, expose it to the pervasive and random yeast in the air around us and if it doesn't mold over first you'll probably get something in the 8% ABV range before the alcohol waste product inhibits the yeasts' metabolism. With some deliberate cultivation, which humans thousands of years ago figured out, you can selectively breed the yeast slurry in your vessels up to a 11-14% alcohol tolerance. They didn't know the science but they knew the process.

Leave it exposed and you'll eventually get a random acetobacter infection and ta-da, now you have a vat of vinegar. Which would still have utility, but is much less fun.

The 1-2% alcohol drinks were called "small beers" and were the second runnings off the grain that was used to make full-strength beers or barleywines. Humans liked variety.

And people drank water, not just alcoholic small beers, there are depictions of water filters dating back at least 3500 years, and a long history of water purification techniques, and a ton of ancient wells which were dug.

My understanding is that Anthropic doesn't really have a lot of say in the matter of keeping their models out of government hands. The government can do pretty much anything they want with a domestic company in the name of national security. They might first choose to be polite about it with an open request and payment, then they might choose to infiltrate and extract by any means necessary (the NSA has a literal army of hackers on hand to point at the task) and Anthropic may not even know about it leaking, either way they do what they want and serve a national security letter to silence the event.

Ultimately they can just offer something Dario can't refuse. Every one of these "oops, our models are just so powerful our agents hacked something by mistake" PR fluff pieces is admission to federal crimes and they are leverage they're giving the government to exert control over them. Refusal means prison.

This is an article and author that is tailored for AI curmudgeons or an AI rejectionist to self-assure themselves that their worldview is the only correct one. Facts seem to be a nuisance at best here for the author, and helpful phrases like "garbage", "vile", "human toll", "psychopaths" undermine all chances or credibility for a reasonable take or conversation.