Natalie Cainaru

The problem, as far as the ethics side of this goes, is of course that there is no One Cool Trick that can create utopia. No one could ever think it, no computer could ever calculate it.

There is only the messy business of trying to do the best we can, for who we can, wherever we find ourselves.

So are these people liars or fools, when they say things like that? Probably both.

So yes, it is silly when they say “I’m teaching the computer to be ethical.” Yes, of course it’s not true that Claude can think, let alone be ethical (or unethical).

But their view of the world is internally consistent in the way all cults are.

Of course, this project is junk from the start. The understanding of the human, of rationality, the utilitarianism, and the view of AI as a self-aware intelligence is all garbage.

And hence, again, AI comes in - here in EA-land as a thing which could calculate what to do to maximize utils. And again, these people are driven by the terror that they might perform an action that is less than maximally beneficial by their own lights.

(This, of course, is how EA connects up to that idiotic “what if AI were god” thought experiment. It should be unsurprising, for example, that Askell has a paper on Pascal’s Wager. We are talking, at the largest level, about the same group of people as those who feel a need to create AI god)

Hence the drift EA has undergone to longtermism. Their utilitarianism means that the present, and presently existing people, don’t matter (because people don’t matter). So they’re not concerned with “trivialities” like genocide in Gaza or ICE in Minneapolis. No, they care about asteroids and Skynet.

On the ethics side of things, these EA people are essentially searching for a silver bullet that can solve the world

(Because they view ethics as an input/output program, and view humans as bearers-of-value rather than as people)

Or rather, the fact that we can be misled is terrifying for them. Because all sorts of things can affect your judgment. Hence AI: a reasoning system (that, again, “reasons” in the same way they believe we “reason”) only without all the messiness of being human

Also, as a pet peeve - "artificial" intelligence literally means something that APPEARS to be intelligent, but isn't ...

So if we do actually create a true machine intelligence, it should be called "SI", "synthetic intelligence".

But this comment highlights an important piece of the puzzle as well: If you think this reasoning is what we do, and ought to do, and this is the foundation of your ethical system, the fact that this is not how we are is a problem

I think this comment is correct: these techno-utopian EA weirdos do think that humans are data in/data out machines. And that’s tied up with their peculiarly awful kind of utilitarianism

They believe that human thinking (“Reason”) is like this, and believe that’s how we ought to approach life

People are (rightly, of course) criticizing this for pretending that AI is self-aware, but I continue to worry that people are missing the bigger picture weirdness that gives rise to that. A thread follows, mainly focusing on the ethics side of things.

Did you see scientists crossed DNA of a cheetah with a crab? It went sideways real fast.

Because ChatGPT is going to be selling CSAM for profit by the end of the year.

Opinion | Affordability and the ‘Epstein Class’ Will Define American Politics

“When Citizens United was decided in 2010, billionaires had spent $18 million on the 2000 election, $13M 2004 and $16M in 2008. Then came the deluge. In 2012 it was $231M, and nearly doubled again in next three election cycles —to $682M in 2016, $1.2 billion in 2020 and $2.6 billion in 2024.” Gift:

DHS shot her and called her a “terrorist.” New videos show something different. "It’s time to get aggressive," a Border Patrol agent said shortly before Marimar Martinez was shot five times.

Like in the killings of Renee Good and Alex Pretti in Minneapolis, DHS might have gotten away with its false claims about Martinez were it not for the video evidence that contradicted its account.

Not much of a lesson from my personal experience, other than: it's okay to say no to potentially sketchy things. I've said no to a few of them; some probably harmless, some doubtless bad. Your integrity as a person isn't worth a few bucks or a few connections.

Gross! Also, it’s chilling to see how much these two people drove the emergence of the far-right across the world from behind the scenes over the past decade or so.

we need to talk about that Ring Super Bowl ad

Graffiti on a boarded-up window reads "NO BARKING" with the top of a car visible in the foreground.

i think they’re being overly optimistic regarding the canine literacy rate

One thing the last few years have taught us, and the Epstein files have only reinforced, is that the people behind Silicon Valley have no respect for the idea of consent.

Keep calling American Oligarchs the Epstein Class. Make it stick

The connections among Epstein, AI oligarchs, cognitive scientists, eugenicists, misogynists, rapists, and outright fascists are so strong it's stunning that they are not outlined in every account of AI development and deployment.

If you are bullish on AI you might want to check yourself.

Electronic Waste Graveyard Immortalizes Dead Electronics Hackaday Article

Electronic Waste Graveyard Immortalizes Dead Electronics

We gotta come up with a better system than “everything rests on whether these twelve billionnaires are nice”

ohhh ‘democracy dies in darkness’ was aspirational

Join The Onion Don’t just read the news. Feel it. Be among the first to feel the news.

Fuck Jeff Bezos. Subscribe to The Onion.

membership.theonion.com