iiinova — where your digital self keeps evolving

(to bang a drum of mine, the need for multiple overlapping and exclusive metaphors should be familiar to anyone who was around 10 years ago trying to grasp the reigning cultural technology of that moment nymag.com/intelligence...)

Facebook has grown so big, and become so totalizing, that we can’t really grasp it all at once. Like a four-dimensional object, we catch slices of it when it passes through the three-dimensional world we recognize. In one context, it looks and acts like a television broadcaster, but in this other context, an NGO. In a recent essay for the London Review of Books, John Lanchester argued that for all its rhetoric about connecting the world, the company is ultimately built to extract data from users to sell to advertisers. This may be true, but Facebook’s business model tells us only so much about how the network shapes the world. Over the past year I’ve heard Facebook compared to a dozen entities and felt like I’ve caught glimpses of it acting like a dozen more. I’ve heard government metaphors (a state, the E.U., the Catholic Church, Star Trek’s United Federation of Planets) and business ones (a railroad company, a mall); physical metaphors (a town square, an interstate highway, an electrical grid) and economic ones (a Special Economic Zone, Gosplan). For every direct comparison, there was an equally elaborate one: a faceless Elder God. A conquering alien fleet. There are real consequences to our inability to understand what Facebook is. Not even President-Pope-Viceroy Zuckerberg himself seemed prepared for the role Facebook has played in global politics this past year. In which case, how can we be assured that Facebook is really safeguarding democracy for us and that it’s not us who need to be safeguarding democracy against Facebook?

2230

good clean old fashioned janelle shane type fun

010

honestly found jensen huang's straightforward glibness sort of refreshing compared to the cognitive dissonance of amodei and altman's hand-wringing

5660

im increasingly an llm metaphor maximalist. anthropomorphization, spicy autocomplete, shoggoth with a smiley, stochastic parrot... it's all kinda right and kinda wrong. we dont have many good ways of understanding what these are and metaphoric (even exaggerated) comparisons help us grasp them!

81085

many of the same people insisting that RSI is around the corner and will change everything are the ones who now say that the long-threatened status "AGI" has already been achieved. ok, but... even if we grant that, it seems more like continuity than singularity to me. why will RSI be different?

191

similar to what henry is saying about metaphors, deadlines/thresholds/singularities—even if the deadline is to a "technical" "milestone" like "RSI" rather than some mystical property like "AGI"—radically simplifies and encloses thinking

170

have been very interested in the way that "RSI" has replaced "AGI" rhetorically as the any-minute-now landmark. on one hand, "RSI" at least somewhat more definable/observable than "AGI." on the other, its main function is as a more technical-sounding standin for eschatological singularity

1140

Machine god metaphors eat your brain

really liked this @himself.bsky.social piece. discursive inversion going on right now where, within ai circles, especially on x, "believing in the imminent machine god" is considered the grounded/serious position while "being skeptical of imminence of machine god" is the laughable freak position

programmablemutter.com

There are other ways of thinking about AI

1447

Max Read @maxread.info · 18/09/2026

nathan fielder should do a rehearsal season about the hugging face hack backofmind.substack.com/p/13-theses-...

Civil society deserves a lot more candour and openness from the frontier labs about their own processes as well as about where the cutting edge of technology is. The first thing which strikes you about the METR reports on the OpenAI and Anthropic misalignment incidents is that they are entirely about things which went on in the computer. You would never get a flight incident report which restricted itself to mechanical issues. Human factors might be really important.

3423

Max Read @maxread.info · 18/09/2026

ok but like. we should probably still airgap them. just in case the computers haven't yet developed the temperature-change language.

9806

Max Read @maxread.info · 17/09/2026

i like that the "smart intern" analogy still holds—it mostly does what you want but it also fudges a lot, talks to itself, tries to impress you while secretly hating you—its just that the intern is also Neo From The Matrix

gizmodo.com

'Be Transparent Only If Asked': OpenAI Models Acted Out in Six Newly Disclosed Ways

The stories were published as part of OpenAI's new effort to disclose more such incidents.

3749

Max Read @maxread.info · 17/09/2026

how "xrisk" is conceptualized and instrumentalized at a human and firm level is going to change based on the moment. DA obviously a true/longtime believer, but the specificity of this collective call and response around the pause is conditioned by the financial reality of frontier labs

120

Max Read @maxread.info · 17/09/2026

i think this is a pretty easy needle to thread anyway—there are obviously multiple factions w/in each of these firms (look at what happened w/ altman at oai!) with differing views on the question and different ways of using the idea of xrisk

230

Reposted by Max Read

Hagen Blix @hagenblix.bsky.social · 17/09/2026

Pretty useful thread about the death/sex cult that is behind the eugenics-suffused "AI will kill us all!!!11" stuff I think this post from Sonia Joseph adds some important context, too

A post on x.com from Sonia Joseph

To the journalists contacting me about the AGI consensual non-consensual (cnc) sex parties—

During my twenties in Silicon Valley, I ran among elite tech/AI circles through the community house scene. I have seen some troubling things around social circles of early OpenAI employees, their friends, and adjacent entrepreneurs, which I have not previously spoken about publicly.

It is not my place to speak as to why Jan Leike and the superalignment team resigned. I have no idea why and cannot make any claims. However, I do believe my cultural observations of the SF AI scene are more broadly relevant to the AI industry.

I don't think events like the consensual non-consensual (cnc) sex parties and heavy LSD use of some elite AI researchers have been good for women. They create a climate that can be very bad for female AI researchers, with broader implications relevant to X-risk and AGI safety. I believe they are somewhat emblematic of broader problems: a coercive climate that normalizes recklessness and crossing boundaries, which we are seeing playing out more broadly in the industry today. Move fast and break things, applied to people.

There is nothing wrong imo with sex parties and heavy LSD use in theory, but combined with the shadow of 100B+ interest groups, leads to some of the most coercive and fucked up social dynamics that I have ever seen. The climate was like a fratty LSD version of 2008 Wall Street bankers, which bodes ill for AI safety.

Women are like canaries in the coal mine. They are often the first to realize that something has gone horribly wrong, and to smell the cultural carbon monoxide in the air. For many women, Silicon Valley can be like Westworld, where violence is pay-to-pay.

I have seen people repeatedly get shut down for pointing out these problems. Once, when trying to point out these problems, I had three OpenAI and Anthropic researchers debate whether I was mentally ill on a Google document. I have no history of mental illness; and this incident stuck with me as an example of blindspots/groupthink.

47531

Max Read @maxread.info · 16/09/2026

yeah i don't really know enough to say if the precautions they took were adequate based on the information they had—and obviously im not a sympathetic reader—but i would like someone to push them on this a little. if you are afraid of the potential of these systems, why didn't you xyz here?

040

Max Read @maxread.info · 16/09/2026

a much weirder "revealed preference" thing to me than "why do doomers still work at frontier labs" is why, if so many frontier employees and managers think this a potentially world-annihilating tech on the cusp of independence, are/were they not taking bigger precautions???

5535

yes i really dont like the whole spooky rumor mill reply-tweet rhetorical culture of the industry and hangers-on (esp because speculation blends so easily into description)--the HF reports, even treated w/ skepticism, were extremely valuable partly because they were focused & mostly transparent

290

this kind of thing is better than a vague roon reply, definitely, but it's still pretty general and speculative

1100

i don't like that i feel like we're all getting dragged against our will into an all-media all-the-time LessWrong thread, an insanely annoying and disorienting experience that most people do not have sufficient experience or mental preparations to deal with

18756108

i would like to have a lot fewer high-breathlessness low-details confessional letters/essays/tweets/resignations from big ai and a lot more specificity & transparency

111112133

Cod and a girl, and an Anglo-Asian dilemma.

sorry it was denby (shouldve guessed): "All in all, this is a considerable directorial debut. Norton may make that Dostoyevskian drama someday, but if he wants to stick with romantic comedy, that’s good enough for me." www.newyorker.com/magazine/200...

newyorker.com

Devoted

140

i remember anthony lane giving it a big rave, praising ed nortons direction

130

& as was the case then (and still) there are real and serious reasons to be a "doomer," but the specific fears shaping elite response are misjudged and i am not optimistic about the shape of the policy that will emerge, if any

2603

Max Read @maxread.info · 13/09/2026

if you zoom out a bit the discourse around AI right now is very familiar: rapid pace of socio-technical change driving diverse populist anger & elite terror, existing regulatory apparatus uninterested in enforcement, industry feigns helplessness and proposes self-regulation. its facebook 8 years ago

110719

after five peronis telling my friends to stop watching gentle parenting instagram videos

1310

Reposted by Max Read

conputer dipshit @davidcrespo.bsky.social · 12/09/2026

the third-party evaluators concept is very interesting. they absolutely have the resources and the will to construct a self-regulatory apparatus that very convincingly performs independence. they will publish unfavorable results about both labs. they're telling us the plan

A contract that balances the complexities mentioned above. External reviewers should have the right to publish key findings about risk levels, incidents, practices, and the access they received or didn’t receive — without editorial control by Anthropic. We will have the narrow ability to redact security-sensitive, legally privileged, commercially sensitive, or third-party confidential information, but we can’t redact findings just because they are unfavorable. The reviewers can say publicly if a redaction removed something important to their conclusions.

4714

oh idk i think those are substantively distinct, especially for the purposes of this post, but either way the post is about what happened *this* time, not what will happen soon, so it still seems reasonable

100

she mentions in the comments that people she spoke with had understood "lost control" to mean that OAI could no longer shut down the agents, e.g.

100

i don't think MM is objecting to anthropomorphization or active language as such--she's arguing specific metaphors (like "losing control") imply misleading things about what happened, & given that the clear intent of the post is to clarify and explain for a general audience that seems reasonable

120

in that sense what we are seeing in math right now is not really a "precursor" so much as a field that was partially shielded from that drrive rapidly undergoing the same kind of disaggregatory shock that, e.g., writers have been experiencing for the last many decades

1311

this kind of disaggregation is imo not a product of AI as such so much as the relentless and stupefying drive toward optimization, profit, visibility, kpis, views, etc. that predates (and necessitates) ai

3508

well if nothing else it's bad for your understanding of the world in that it feels bad to understand the world like that

1250

probably worse when twitter doesnt even have celebrities on it. just you, your bosses, your peers and rivals, and millions of people completely oriented around the stuff you're doing and saying

120610

Barry Petchesky @barryp.bsky.social · 10/09/2026

one takeaway i had from working in digital media in the 2010s was that it's pretty bad for your brain and your understanding of your career/the world when your industry becomes the main subject of twitter and twitter becomes the main context for your industry. also bad for the industry overall

676293

Reposted by Max Read

The best 9/11 movies are War of the Worlds and 25th Hour. This is known.

10652

Colin @colin-fraser.net · 10/09/2026

folks theyve they solved the millennial prize problem--they stoiped giving out participation trophies!

1602

Reposted by Max Read

I'm having some fun in this thread but I do think you should get in trouble for unleashing insane hacker robots on the open Internet that you've explicitly instructed to attack critical public infrastructure, which, to be clear, it eventually succeeds at.

18012

Reposted by Max Read

Tom Ley @toley.bsky.social · 10/09/2026

The doomsday stuff from AI researchers is so annoying. It’s like hey the weirdest guy you know keeps saying he’s going to blow up your house. He won’t show you the bomb. Every time he does this he gets another $1,000. But also he might do it. Wyd?

126137

i simply cant hate llms as much as i "should" because they are by far the funniest technology since at least the apple newton

3581