Speaking of movies about high-profile silicon valley grifters, what the fuck is this??

At this point given the absurdity of this combination of people I feel like the reality will likely be disappointing.

It also causes me no end of rage as a parent that this message is coming from this administration in particular since their policies are at least half the reasons why we aren’t getting out in the world more. Like, my family has queer and trans people in it and the administration have explicitly aligned themselves with and empowered people for whom that is a reason we shouldn’t be a family. The cuts they’ve made to childhood vaccine access and the way they’ve boosted antivax bullshit means that my immunocompromised wife can rely that much less on the population’s herd immunity. The ongoing economic fuckery and cost of living crisis that they’ve inflamed through their pointless goddamn war in Iran and general incompetence have greatly increased the costs of getting to anywhere in ‘real life’, especially since the communities we can trust in light of the above are pretty geographically scattered compared to like 15 years ago. But yeah, the problem is the screens somehow.

Hell, the inscrutable words used to be things like “queer” or “lesbian”.

What I’m saying is that in like 20 years we’re going to hear about the actual emperor’s (of where? Who knows. Probably like Michigan) preferred danmei series and we’re going to like it.

I guess maybe the sex cult and the ritual magic make the standard grifting seem credible by comparison? Like, E-Dub and company don’t pass the sniff test but Saltman and friends seem like reasonable adults who can’t all be wrong?

I followed the kinks back. Apparently these are phrases that emerged from the OpenAI “message board” that was used to “coordinate” the huggingface incident. Or rather, these phrases were extracted from the junk data that these processes left strewn around the ecosystem because OpenAI apparently has a more laid-back attitude to the principle of least permission than Jurassic Park.

There’s a bizarre tendency in a lot of technocrats to basically reject the notion that people in the future will have the power to make different choices. Like, it’s not enough to put together a policy that will solve the immediate problem, no matter how pressing, because if you extend the resulting trend line infinitely it might result in a suboptimal outcome. Combined with the TESCREAL project’s long-termism and obsession with personal immortality it feels like the link to far-right paternal totalitarianism is actually pretty straightforward.

If I were inclined to research and source properly I feel like there could be some meat there that I don’t think Emile Torres and his colleagues have gotten all the way into yet.

Maybe some kind of diagram that showed the links visually, color-coded to illustrate some more details about the connection. Add some photographs for visual appeal.

I don’t think any of the words on that shirt are in the Bible.

Nonzero chance AI kills us before we reach the end of this probably-years-long-process, but we plant trees in whose shade we may never sit.

You know, I didn’t think it would be necessary to discuss the meaning of that line in a space where people ostensibly take long-termism seriously.

AI Use: LLM use is permitted, but discouraged; we remind participants that LLM use tends to reduce originality and writing quality, and the flaws are especially obvious when LLM outputs are read as a group—unskillful use of LLMs will reduce the odds of being the best entry. We request disclosure of AI use.

Lol. Lmao.

I guess given the demographics of the diehard Bernie bros this isn’t necessarily a surprising turn, but it’s still disappointing.

Also, hat tip to Dr. Rasmussen for cogently articulating the frustration of the hour. Why in the hell are we talking about whether the acausal robot God is going to endanger our ability to trade galaxies like Pokemon cards when so much of the world is both literally and figuratively on fucking fire?

Tag yourself I’m “(I suspect)”

I don’t think it’s a very novel insight to point out how this pattern is pretty typical of high control groups, though I guess most are much more explicit about the in-group/out-group distinction there. Our very good friends seem less inclined to treat everyone outside the circle like they’re part of a grand anti-tescreal conspiracy, at least.

This may be a little bit off topic, but since it sits at the intersection of academic freedom, race science, and plagiarism I honestly can’t imagine our very good friends staying quiet on the matter for long.

Jason Arday, youngest black professor in Cambridge history, dead at 41 after plagiarism debacle

It looks like the initial accusations here came from Nathan Cofnas, a notorious racist shitheel. I am not qualified to speak on the actual validity of the plagiarism accusations, but suffice to say that my priors are very low. Cofnas’ current employer seems to agree at least that the optics are pretty fucking bad, and he’s been given a teaching ban and presumably a private request to stay out of the spotlight for a while.

It seems like some of the usual suspects are making noise about that from the academic freedom angle, but I think the more salient criticism is that it’s get another example of academic culture being unable and/or unwilling to handle people acting in bad faith. Like, even if the plagiarism accusations are completely substantiated, I don’t think it’s unreasonable to assume that Cofnas was looking to harass a prominent black professor at a well-regarded university and that the low-effort “the AI app says this is plagiarized” attack was simply an expedient means to do so. Race “science” is still a political project masquerading as academic inquiry. Or rather, the line between “impartial” academic inquiry and political activity is far less coherent than advertised and failing to acknowledge this will only empower assholes like Cofnas to continue their bullshit.

As is so often the case, I honestly don’t know what’s supposed to be different about our response to gengineered superplagues if they were made by an AI instead of the CIA or whoever else is in on that nonsense.

Always love reading Iris’ work. I’m already finding ways to abuse the concept of the person-tool gestalt for dumb jokes that I have to over explain to my wife so I’m getting my money’s worth to be sure.

Also I definitely have actual fond memories of randomly running into people in Guild Wars 2. The of genericizing/instancing more of the world interactions (like loot and quests, for example) and using mega servers to dynamically scale resource usage and player allocation meant that you were very unlikely to organically run into the same people often enough to really make even a parasocial connection, but it dramatically improved the experience with the overall community.

Objection! Gas Town isn’t junk, it’s a memetic weapon trying to convince anyone with actual engineering sensibilities to abandon the industry and move to their cabin in the woods to become professional manifesto writers, thus clearing the path for AI inevitability.

Of course, the irony is that in their desire to revolt against the violence being forced upon them they merely created an incredibly efficient means of achievement farming for those with the stomach for it.

It’s also fascinating to see them use nuclear vs renewable energy as their case study given that renewable energy has been having a hell of a time properly replacing fossil fuels even though everyone mostly agrees it’s a good idea. Like, whatever hypot rrtherrrrr see drdrr try r the t6 ty to r the rthetical rradvantagre they got from their opposition to the nuclear boogr the rrrrrreyman (their idiotic framing, not mine), renewables havr guy off bute still been facing some for trightand political headwinds that are only being magnified by the unceasing growth in demand for electrical power (including but not exclusively driven by AI compute).

But if I don’t obsessively optimize my career in ways that prevent me from making meaningful human connections, how am I supposed to keep up with the relentless march of AI??? Thank God I have AI to help pick up the slack.

I wouldn’t put a multiplier on them, except for 1.2x to 1.25x.

I’m pretty sure it’s still a multiplier if you’re not using whole numbers.

I’d be curious to see how they defined “harmful” in their test, since it looks like Claude code’s technically-not-yolo mode is looking for “actions that are irreversible, destructive, or targeted outside of your environment” which covers a lot of normal actions, especially depending on how you define which resources are in “your environment”.

And yes, I’m sure that the fact that the flow of operations they’re outline will massively increase the number of times that the system automatically retries things, this increasing token usage, is going to be fully accounted for when they review the financials on how many tokens people are paying for.