I’m just commenting on the book. I find YouTube videos pretty insufferable. I guess it’s a tangent.

I’ve listened to a couple interviews with the author about this book, and I have not found them persuasive. I can accept that there’s a possibility that artificial super intelligence (ASI) could occur soonish, and is likely to occur eventually. I can accept that such an ASI could choose to do something that kills everyone, and that it would be extremely difficult to stop it.

The two other arguments necessary for the title claim, I see no reason to accept. First that any ASI must necessarily choose to kill everyone. The paper clip scenario is the basic shape of the arguments presented. I think it’s probably impossible to predict what an ASI would want, and very unlikely that it would be so simple minded as to convert the solar system into paper clips. It’s a weird proposal that an ASI must be both incomprehensibly capable and simultaneously brainless.

Second that the alignment problem can not be solved before the super intelligence problem with current trajectories. Again, this may be true, but I do not think it’s a given that the current AI techniques are sufficient for human-level, let alone super-human intelligence.

Overall, the problem is that the author argues that the risk is a certainty. I don’t know what the real risk is, but I do not believe it is 100%. Perhaps it’s a rhetorical concession, an overstatement to scare people into accepting his proposals. Whatever the reason, I’m sympathetic to the actual proposals; that we need better monitoring and safety controls on AI research and hardware, including a moratorium if necessary. The risk isn’t 100% but it’s not 0% either.

Pretty sure they’re typically publicly owned. Maybe some places lease them. Couldn’t find a national survey, but here’s at least one example of a county that bought some machines and a service contract.

https://fm.kuac.org/elections/2025-03-10/assembly-fails-voting-machine-contract-may-force-change-to-hand-counting-ballots

Maybe a car fleet is a good example. Ford designs and builds the cars. Counties buy them, and often buy service and maintenance contracts to keep them running. The counties still own the cars.

I suppose counties could receive the source code, have it audited, and then compile and load it themselves.

I thought about this for a second, and I don’t actually think being open source would do any good. It’s not like we can compile and run our own voting booths. There’s no way to know what’s actually running in the machine at your polling place.

And voting machines are publicly owned, but perhaps you meant designed and manufactured by the government?

Exactly. They’re trying to scare us off. A little courage now may spare us the need for really scary things later.

Do you have a reference on graphene being affected by the app signing? I tried to find an answer earlier, and came up with nothing.

It would be interesting to learn something about the demographics on Lemmy.

I usually liken the bad vibes on Lemmy to being stuck with a bunch of cynical teenagers. Nothing is ever good enough, nothing good can happen. They know this with absolute certainty.

I am also probably older than average here.

They do, but I’m a little surprised by how well they’ve positioned themselves on this one. It seems to me that the most likely scenario is that the Republicans will give nothing on principal, the shutdown will go until November when the premiums increase, and the country will see that the Republicans would rather close the government for two months than spare them a doubling or tripling of their healthcare costs.

And all the while Trump trashes the government in an attempt to retaliate, without really understanding that the government provides services that people, his voters included, depend on. I’m not sure, “the Democrats made me do it,” will save him with anyone other than his cult members.

I am cautiously optimistic.

Rage bait?

I try to stay skeptical about conspiracy theories, but I have yet to see an explanation for why this guy had so much money and connections. To talk about “science” and “money markets”? At Bill Clinton’s request? None of that makes sense to me.

How did it never occur to me to ask where bee’s wax comes from?

Looking forward to more vindictive prosecution findings.

Recognizing that the physical can affect the mental, and vice versa, isn’t really the end of the dualism argument. Dualists have incorporated that simple observation from the beginning.

From your quote, the key word is “purely.” Is consciousness purely physical, or is some other substance involved, that’s the question.

You can take either side of the argument, but physical-mental interactions only suggest that mental phenomena are not purely separate. It does not indicate that there are no non-physical elements of consciousness. In other words, that mental states are purely physical.

https://en.wikipedia.org/wiki/Mind–body_dualism

If you want to read through some of the arguments for and against.

Are you objecting to him being described as a refugee?

Is this another one of those where Steam Deck gets lumped under Arch?

In 1922?

Last I heard Trump was cold calling world leaders to beg for Nobel peace prize nominations.

I’m not really qualified to evaluate the merits here, but as a science-interested layman, I’d be glad to see an alternative to dark matter and energy. Setting aside the technical arguments, the dark matter and energy approach smells like questioning the observations when your theory doesn’t match observations.

I skimmed the paper for testable predictions, and nothing stood out to me. Fitting existing observations is a good place to start, but if the only prediction is that nobody will find dark matter or energy, things may remain undecided forever.

I thought the healthcare tax credits were too technical to grab popular support, and turn the shutdown against Republicans. Maybe I was wrong.

A counterpoint would be to ask which platforms digg and reddit began as clones of. Seems they were pretty unique and yet exploded almost from the beginning. Snapchat? Vine? They were both pretty unique.

To OPs point, basically all of the fediverse apps are clones, which aside from the federation element, don’t add anything to the formula they are cloning. Even if you prefer the incremental strategy, where things are basically the same with a few new features, it would be hard to argue the fediverse apps even meet that bar. To the average user, federation is a technical issue they’d rather not be bothered with.

So I’m inclined to agree that this first wave of open source, federated social platforms have ended up, in terms of social features, pretty uninnovative. But before I sound too critical, I appreciate the work these app builders have put in, and clearly use the apps myself.

It may be a question of project scope. If what you aim to do is liberate yourself and your fellow nerds from corporate platforms, the clones suffice. If, perhaps, your aim is to liberate everyone, you’ll need innovation in both the backend, and the social features to draw in everyone else.

Caveat - I’ve only really used Mastodon and Lemmy. Perhaps others are different.