Arvind Venkataramani is letting go of conditioning

Here's our field report from running a ritual design studio at Burning Man this year: link.medium.com/a5vz7BQDu0 Curious about ritual design? Check out our project and toolkit at ritualdesign.net

@STS_News Oh yes and this is in continuance of our thread earlier about the lack of data on successful use that doesn’t simply parrot marketing and hype points.

@_arvind The psychosis stuff doesn't surprise me at all. People are kind of dumb and tend to project personhood and intelligence onto these tools when it's not there. (The Turing Test sucks, it turns out.) But I'm not sure this is a reason not to use the tools wisely.

I'm writing a post called "I Just Don't Buy Ethical Arguments Against AI Use," and I feel like this might be a useful place to start putting down ideas. In this thread, I will consider how the term "cognitive offloading" has come to be used - often with implied moral judgment.

@STS_News Aaand… here’s a study on “psychosis” that touches on these themes: arxiv.org/abs/2603.16567

@STS_News Indeed. Something seems qualitatively different in this one because of its apparent anthropomorphic superpowers (for consumption) and its alignment with cybernetic science-fantasies (for production). It’s rare for tech to so explicitly and accessibly refigures “human” IMO/IME.

@NinaDSchick @tunguz This is an interesting claim. Do you have a study link backing it up?

@tunguz So the ‘human in the loop’ is actually the drag. Beyond bureaucracy, even on accuracy. Saw that in early clinical diagnostic data suggesting that AI acting without a clinician is the most effective — more than a clinician with AI. (Least effective being a clinician without AI.)

If you’ve ever had to deal with bureaucracy of any kind you’d realize how far far away we are from even having the basic technologically competent institutions that operate anywhere near the logical/deterministic bare minimum. AI systems will not change that. If AI is to have a TRULY transformative impact on human society, we we will either have to 1) fix and upgrade those processes (a massive undertaking) OR (far more likely) 2) have the AI work around them (essentially handing over all the power to the AI systems)

@razibkhan That’s fascinating, I grew up around Muslims from various walks of life in India and never once would it have occurred to me to code a sari with a religious identity. It always seemed to have more to do with geography: the more north and west of India, the less sari use.

@deepfates “ Agents can do anything you can do on a computer” No this is only believable if you do not have any understanding of what expertise looks like. There is no way to universally achieve quality via quantity; the damage won’t come from “AI” but from using it as an excuse.

You might think the "agents" thing is just coming for software engineers. Yeah, agents write code, code and code sells a bunch of tokens, But most people's work isn't code, it's memos or decks or whatever. Why this is false: Agents can do anything you can do on a computer, and they do it by spending output tokens to write code. The number of keypresses used by a consultant to do a task is not a good measurement of the number of tokens an agent would use. For example: one "deep research" report might be 20 pages of output tokens. But it also might have required more than 20 pages of output tokens to do all the searches, fetches, PDF parsing and interim summaries that you never even see as the user. It also had to input all the tokens of every document it read in searching — likely more than 20 pages, since the point of the report is to collect and summarize this information. So now we're at 3x tokens for the final output. That one report is so cheap, and so fast, then now you can do more research than ever. This is valuable! If your business relies on having good information about the world, you can probably find a way to make more money by doing 3 deep research reports and then synthesizing them. More tokens! Now you've kicked off three deep research reports you deserve a little treat, right? So you fire up your browser agent and tell it go find me some nice linen shirts for summer in my size. Open them in tabs so I can look through. Well your browser agent has to interact with the browser using some kind of tool and you know what that tool is? Code, baby. Tokens. And the tokens are so cheap. You got to understand. We're spending a lot in the aggregate, but in the moment it is "spend a nickel to for 10 minutes of being literally Superman". Like yes I'll just keep spending nickels actually. I will never stop being Superman at that price. All knowledge workers will feel this. A lot of you already do, you're just hiding it from your boss so you can have more free time while "working from home". And maybe it's better to protect yourselves from Jevons as long as possible, because once you get the bug it's hard to stop. You realize that you could be creating all of the businesses and projects and art you ever wanted and all you've got to do is put your instructions in the right order and put the nickels in the bag. I would happily bet against Anthropic's revenue spike being a brief "sugar high". So would most capital allocators! That is because they have already seen that software can eat the world. White collar knowledge work fundamentally changes in the face of agent economics and entirely new forms of knowledge production? It's happened already in finance: high frequency trading. Now it's happening in tech: high frequency software. Then we will have high frequency science, high frequency governance, high frequency engineering, high frequency medicine and high frequency law. Human society is about to be absolutely DDOSed by information at all levels of the stack. Our civilization was never meant to handle this many tokens. If anything can be done on a computer it will be turned into tokens instead of human actions and it will happen faster and in parallel. This stuff works, it is real, it is getting better. It is going to hit economically and socially this year and nobody is ready and I think it is important to start taking it seriously, instead of finding ever more arbitrary reasons to remain in denial.

@provisionalidea I am begging them to achieve at least a bare minimum of dorkage for the public good

@_arvind unfortunately for us, economic and social impacts teams of big labs are a marketing expense and not a place for methods dorks to do real innovation

interesting tool, always love people playing around with new methods I *will* say, judging by the raw transcripts, that it’s more like a mix of junior UX and market research interviews than it is a read on adoption and impacts. Lists of usecases aren’t sociological models!

@provisionalidea Based on how they talk about this, I do not believe they* possess the methodological sophistication or the ontological models to have an analysis engine that can possibly drive inquiry in the way that human social scientists do *= all AI qual research tool makers

(more robust and predictive, btw, because iterating on technique becomes cheaper, and also because scale and natural language advancements allow robust surveys to be conducted on the codification of mass soft survey data — which should be baseline here!)

@mykola “You are one of us whether you like it or not!”

@_arvind yeah fair enough that's how I found myself in TPOT there wasn't a name for what I was until suddenly I was surrounded by other "postrationalists" "But I was never a rationalist!" I explain and they just nod and smile.

gosh it's funny when you have a Utopian's heart but a Cousin's soul and a Brillist's mind, but at the end of the day what can you be but a grey hiveless? On a related note if you haven't you should really read the Terra Ignota books they're some of the best scifi ever written.

My act of quixotic optimism this week is watering & fertilizing the moss growing in-between the roofing cracks on my apartment’s balcony so it survives but also thrives during this heat wave. (It survives every year without my help but this year I’ve decided that is not enough)

@STS_News Rotating the perspective a bit, it seems the one definite use GenAI has is revealing many of the deep fault lines and dysfunctions of an overly individualised and hierarchical civilization.

But I think these concerns are really about much bigger and deeper problems in labor politics than they are about this set of tools, and they should be addressed through collective action and better labor laws. Focusing on GenAI use at the individual level here is a distraction.

@mykola The fact that this is happening frequently explains why they’re offering free usage outside core hours as a “promo” but is probably an experiment to see if they can shift usage to manage load.

uh, is claude down for anyone else? Sudden 500 errors in claude code and the desktop chat app is spinning forever without answering?

@STS_News @libshipwreck Splice x.com/_arvind/status…

@STS_News @libshipwreck And that this benefit is laundered as an excuse But also simultaneously likely that LLMs provide certain kinds of support for some people that we did not realise was possible to give or was needed ↓

Stop using AI. However, if you are going to use it (and to say it again, you shouldn’t) you should say “thank you,” because that way when the robots take over they’ll remember you were nice to them and spare you.

@STS_News @libshipwreck Which then produces the question of: what framework animates that decision-making and why and where does it come from and all the other sociological and historical questions that come from that

@STS_News @libshipwreck Which means we would need to differentiate the argument of using Gen AI is bad for the user from using Gen AI is bad for the ecosystem/public/commons, …knowing that some people will choose to respond by ignoring the collective harms