Only the magic eight ball has been rigged with sides reading:

• Signs point to yes

• It is decidedly so

• Absolutely. You’re so smart

• Maybe. Good question!

• There are strong reasons to think so

• Lots of people are saying it

• I can see why you’d ask that

• There isn’t a strong consensus either way

Is that what “g” has stood for all this time?

But despite well-documented claims to genius IQs, somehow the billionaire set ends up not on the chopping block.

To be fair, while I’m not familiar with the discourse in China I know a lot people consider (rightly) “alignment” as a framing to be a red flag for cranks and rats. It’s not that surprising that this attitude hasn’t been getting much recognition when the marketing departments of ai companies has been more engaged on that subject than serious academics.

Okay but hear me out: if they do restrict it to only Russian training data then it’s gonna end up having a small enough set to fail in truly comical ways.

You can see also his rabbit counterpart from Zootopia in a princess dress in the background of the first picture!

I’m sorry, I think I need to believe that this is taking the piss in order to be able to function. It can’t be real (It’'s definitely real).

Oh God I read their FAQ and it looks like the whole concept is to gamify smoking weed because if there’s one problem with weed it’s that it’s not addictive enough on its own. I mean the actual concept is to try and smash enough hip tech buzzwords together to extract some amount of the dwindling venture capital continuing to slosh around the valley, but if it actually happens the thing it’s going to do is take all the addictionware tactics that app developers have developed and bring them to bear on promoting drug use.

Yeah. I mean, I’m going to assume that he’s not specifically referring to LLCs and that, say General Partnerships and Sole Proprietorship aren’t actually better, mostly because they absolutely share the same problems. But that’s a whole lot of digital ink spilled about how important it is to create “digital public goods” - and one assumes public goods more generally, given our ongoing inability to subsist on digital food - without any real thought put to how such systems should be organized, controlled, protected, etc. Like, if he stopped trying to wave “digital innovation” around like a magic wand he’d basically have to be a communist or anarchist of some flavor. But he’s still sufficiently in the Silicon Valley Milieu where lefty politics are too cringe to admit, so he’s stuc having identified some real problems and having no realistic starting point for solving them

Dramatic fascistic “RETVRN” language and focus on aesthetics aside, my wife and I actually dug into some of this lighting quality stuff a while ago and while our very good friend here does a poor job explaining it there is a definite difference in normal LEDs vs incandescent or natural light. The LED spectra is fascinating - big spikes at a couple wavelengths and nothing in between. In my experience with switching to the fancier high-CRI LEDs the difference is pretty minimal. Feels like a possible case where you don’t notice it, but your brain does. For my wife it seems to have helped reduce the incidence and severity of her crippling migraines, which is obviously more impactful. I don’t think I’d say it beautified the space or brought us back to the halcyon days of our glorious past, but that’s been huge for us all the same. The plural of anecdote is not cliche, but there’s not nothing here.

I can’t imagine the Mouse being happy about this cameo.

Okay much as I’m angry and want to I’m resisting the urge to go point-by-point until I have more time. But I also want to point out that in form it seems like The Beigeness has really caught on as a writing style. Like, we have 22(!) individual points, each of which gestures vaguely at the kind of militant interventionist white nationalist technocracy that could conceivably power the unholy chimera of a silicon valley tech giant and a murderous beltway defense contractor. But unless the book does so more openly, they avoid clearly stating the actual thesis. It’s not really surprising, just interesting to note the pattern spreading from Rat spaces into the broader right wing.

On one hand, I appreciate their acknowledgement that legitimacy matters to a government’s ability to govern. While the talk about the king as a figure tied to a broader structure that creates obligations and requirements just as strongly as it does power and privilege isn’t entirely historically accurate it’s at least less absurd than Yarvin’s notion of the dictator as a kind of unmoved mover - someone with both absolute power and absolute discretion to do what they want with it.

At the same time, if you follow that chain of thought to it’s actual conclusions you end up with some kind of radical democracy. Like, legitimacy is just a way to ask the question of why anyone should bother to do what the guy calling himself king says. Historically speaking this often boils down to trying to judge how credible the threat of violence is should you refuse. If the king isn’t going to be around in a week due to an ongoing succession crisis then there’s no point in getting ready to pay his taxes next month, essentially. But if we reframe the question another answer becomes available: why should people consent to be governed? And the democratic answer is that the government represents their interests and is trying to organize and take actions they support. Government by consent of the governed is a descriptive statement about how governments operate, not a normative one about how they should. Once you account for the extra costs and consequences of needing to manufacture consent through violence and repression the supposed efficiency of dictatorship evaporates.

Big “I used to, but I still do too” moment there, though.

I’m glad someone else was able to coherently discuss how ass-backwards Saltman’s response has been. Like, if anything the fact that he responds to this moment by talking up the importance of democracy over emerging technologies should just be evidence before some distant future revolutionary tribunal that he knows his company is literally Sauron (okay, maybe more the Witch-King of Angmar than Sauron) and doesn’t care because he wants to be the one wearing the ring at the end of the day.

From the second post:

A seasoned security leader would never build a defensive program and then measure offensive capability only, making remediation a second-class story. That is the kind of dog and pony show that any good security initiative would slam the door on. Or it’s like a surgeon telling you they have an even sharper scalpel to cut you deeper and faster. Yeah, so then what?

Dark and paranoid thought: given that Anthropic very recently ran into issues with their defense contracts, are they playing up their offensive capabilities targeting a notoriously tech- and security-illiterate political establishment to try and force their way back into those sweet government contracts as an impossible-to-ignore offensive tool? I mean we’ve talked about how the cash burn rate for all these companies is sufficiently absurd that it’s going to take something truly crazy to turn these companies self-sustaining before the world runs out of investor money, and military and intelligence budgets are notorious for dragging ludicrous amounts of public money into a dark alley where nobody can see what’s happening to it.

I mean, I’m not even sure that’s true

Now that’s not fair. It’s based on a third-derivative of Advanced Homework: The Game.

Psychoanalysis really does seem to push the most obnoxious boundary in academic language. On one hand, it is legitimately valuable to create a specific framework to enable experts to talk about technical elements of the field. It reminds me of the old IT rant about users who think “turn on the computer” means “turn the screen on, no need to touch the actual computer part”. But at the extreme it creates opacity for its own sake and makes it hard for people who haven’t devoted their careers to the field to understand what’s being done. Particularly in a medical or psychiatric field where the patient is by definition in a lower-information group than the person treating them, this amounts to making it hard for the patient to understand (and therefore consent) to what is being done to them. I am by no means immune to the simple pleasure of knowing something that other people don’t, especially when the outside world reaffirms the value of that knowledge, and there is definitely a place for the specificity that this kind of jargon enables, but psychoanalysis seems to consistently stretch it too far.

To distract us from the ongoing cycle of violence and discourse about violence that neither cracks down or addresses it’s causes, may I offer the fruit of today’s YouTube rabbit hole:

AI isn’t the future. It’s medieval alchemy.

This feels somehow tied to the whole “agentic” thing I’ve ranged about previously. Like, individual acts of violence are strictly destructive because the people doing it aren’t sufficiently “agentic” to change things, even though American history is full of cases where (usually racist) vigilante violence had a huge impact on people’s decision-making. But when the government does it it’s different because people in government got there by proving their agency and ability to actually impact the world. Like, it feels almost like he’s offended that the NPCs might try and do something as drastic as killing someone without GM permission.

Meanwhile in reality, people legitimately do feel like they don’t have a lot of options to protect themselves from the real harms this industry is doing, to say nothing of the people who buy his line about the oncoming class-K end-of-life scenario. Anger is an appropriate response to the circumstances we find ourselves in, and in a nation that has been quietly cultivating a culture of heroic violence for decades we shouldn’t be surprised to see people trying to inflict that fear and rage upon the outside world.