iiinova — where your digital self keeps evolving

Reposted by Nathalie Maréchal, PhD

Carl T. Bergstrom @carlbergstrom.com · 9m

Columnist Stephen Rattner probably got screwed over by the headline writer, but I'm seriously floored at how poor a job these six charts do at showing that we need AI.

nytimes.com

Opinion | Six Charts That Show Just How Much We Need A.I. (Gift Article)

All the worry about artificial intelligence should not obscure this key point: If we don’t develop it, we’ll be a lot worse off.

1394

Nathalie Maréchal, PhD @nathaliemarechal.net · 13m

Infuriating thread. Thank you, @aylataylor.bsky.social for the report-out!

040

Nathalie Maréchal, PhD @nathaliemarechal.net · 1h

Pour les francophones: le journaliste indépendant français Grégory Aimar a traduit certains extraits de mon papier récent sur le culte de l’IA et les a commentés sur son Substack. Un grand merci! gregoryaimar.substack.com/p/ce-que-les...

gregoryaimar.substack.com

« Ce que les décideurs et le public doivent savoir sur le culte de l’IA »

Prise de conscience en cours de téléchargement chez certains chercheurs.

032

Reposted by Nathalie Maréchal, PhD

Joe Bak-Coleman @jbakcoleman.bsky.social · 02/10/2026

🚨🚨🚨“AI Researcher”🚨🚨🚨 Translates fairly reliably to “person with more coefficient giving money then expertise”

2121

Reposted by Nathalie Maréchal, PhD

Matt Thiessen @mattthiessen.bsky.social · 02/10/2026

your daily reminder that the field of religious studies remains relevant. In fact, it is increasingly relevant.

19937

Reposted by Nathalie Maréchal, PhD

James Felix Black @tft.io · 02/10/2026

The simplest explanation is that they believe what they say.

4275

Reposted by Nathalie Maréchal, PhD

Dominique Baker @bakerdphd.bsky.social · 02/10/2026

Writing something where I will be liberally citing @lilymasonphd.bsky.social's latest and realizing that I can't cite every single thing I would like to because then I'm just reprinting her work. So, I've gotta thread about it instead. muse.jhu.edu/pub/1/articl...

Screenshot of the article website. Title is: Polarization Is Not the Problem. Author is Lilliana Mason. Journal is the Journal of Democracy, Volume 37, Number 4, October 2026. From Johns Hopkins University Press.

8854273

Nathalie Maréchal, PhD @nathaliemarechal.net · 02/10/2026

Thanks for the clarification! I’d love to discuss more but I’m on deadline, and not sure this is the most productive forum for nuanced discussion. Would love to find another time/medium to chat though.

110

Nathalie Maréchal, PhD @nathaliemarechal.net · 02/10/2026

Briefly: You are correct that I do not take AI cult beliefs seriously at all. I take the fact that so many ppl believe in AGI, x-risk etc very seriously, but not the claims themselves. And, good luck trying to change the constitution.

110

Nathalie Maréchal, PhD @nathaliemarechal.net · 02/10/2026

Briefly: I don’t think this is a productive forum for this level of deep, nuanced analysis and discussion. Happy to schedule a call though 🙂

100

Nathalie Maréchal, PhD @nathaliemarechal.net · 02/10/2026

2: because I have yet to see a serious policy proposal that would simultaneously be constitutional, effective, and that would be signed into law anytime soon. There are opportunity costs to tilting at windmills.

100

Nathalie Maréchal, PhD @nathaliemarechal.net · 02/10/2026

2 answers: 1) Because the people pushing the hardest to address “catastrophic” and “existential” risks think addressing the more quotidian (albeit severe) harms is a “donkey sanctuary” ie not “effective”. See this LessWrong post for an example of this reasoning www.lesswrong.com/posts/cr5pyW...

lesswrong.com

AI Safety Acculturation is Neglected — LessWrong

At the local AI safety co-working space, there are ~two kinds of regulars. …

130

Reposted by Nathalie Maréchal, PhD

Hisham Zerriffi @hishamzerriffi.bsky.social · 02/10/2026

Pew tested AI synthetic surveys against actual survey data. Conclusion: “This experiment taught us that, at this time, AI models are not an adequate replacement for traditional polling on topics of broad public importance.” This is an all time bad idea. No need to hedge with “at this time.”

pewresearch.org

Can AI Stand In for Human Survey-Takers? Not Really

Synthetic surveys using digital twins do a poor job estimating the public mood and are especially error-prone for some subgroups, including Republicans and Black adults.

1023077

Nathalie Maréchal, PhD @nathaliemarechal.net · 02/10/2026

I don’t object to that in principle, but it’s not clear to me how it’s possible to write a law that does that without a host of unintended consequences. And that is actually constitutional.

100

Nathalie Maréchal, PhD @nathaliemarechal.net · 02/10/2026

That is not at all what I said. There are MANY reasons to be concerned about “AI” - not least of which is the industry’s willful destruction of US democracy - but Computer God turning us into paperclips isn’t one of them.

001

Nathalie Maréchal, PhD @nathaliemarechal.net · 02/10/2026

You have no idea how badly I needed this eye rinse 🙏🏻

010

Reposted by Nathalie Maréchal, PhD

David M. Perry @lollardfish.bsky.social · 02/10/2026

Mad libs out of control

210515

Nathalie Maréchal, PhD @nathaliemarechal.net · 02/10/2026

To be fair, this is their skeet about the story bsky.app/profile/econ...

111

Nathalie Maréchal, PhD @nathaliemarechal.net · 01/10/2026

Correct. The vision Big Tech is selling to corporate clients and investors is, “replace your workforce with machines! Lower your costs, no one will complain or push back against your BS, and you don’t have to worry about their welfare.”

020

Nathalie Maréchal, PhD @nathaliemarechal.net · 01/10/2026

They’re about as deep as Harry Potter fanfic.

010

Reposted by Nathalie Maréchal, PhD

Lily Mason @lilymasonphd.bsky.social · 01/10/2026

My new JOD piece arguing against the “cure polarization” approach to saving democracy.

10487165

Nathalie Maréchal, PhD @nathaliemarechal.net · 01/10/2026

uh, kind of ? I invite you to read my explainer if you haven’t bsky.app/profile/nath...

010

Nathalie Maréchal, PhD @nathaliemarechal.net · 01/10/2026

That’s exactly it. Even after all the <gesticulates wildly> of the past decade, it’s really hard to process that people really do believe such nutty things. Though is it truly any nuttier than the MAGA worldview ? (No, not really IMHO.)

120

Nathalie Maréchal, PhD @nathaliemarechal.net · 01/10/2026

This is what I’ve been trying to tell y’all about the broligarchs, effective altruists, longtermists and other assorted followers of the cult of AI. They legit believe they are building a computer god and/or a synthetic species that will replace people and conquer the stars. Yes it’s bonkers.

48315

Reposted by Nathalie Maréchal, PhD

Data & Society @datasociety.bsky.social · 01/10/2026

D&S's @mwoluchem.bsky.social describes her team’s research on Pennsylvania’s industrial lineage, the mythmaking that builds on it, and the pushback that defines this moment. “We learned that it takes a lot of work to make an AI factory seem inevitable,” she says. www.youtube.com/watch?v=ZPgA...

youtube.com

[Video] The AI Factory: Data Centers, Power, and Resistance in Late Industrial Pennsylvania

YouTube video by Data & Society Research Institute

1107

Reposted by Nathalie Maréchal, PhD

Ben Winters @benwinters.bsky.social · 29/09/2026

that chatbot doing that unsafe/violative thing is not it 'going rogue' or an 'unfortunate incident of unintended consequences,' it’s the logical and devastating consequences of an unsafe product put out by irresponsible companies that have been conditioned to believe consequences dont exist for them

7789252

Nathalie Maréchal, PhD @nathaliemarechal.net · 01/10/2026

This is why the technofascist project will ultimately fail. Its heart is the desire to replace human complexity, friction and unpredictability with cold computation to avoid ever having the do the hard work of politics.

072

Nathalie Maréchal, PhD @nathaliemarechal.net · 01/10/2026

The fascists really have lost the plot, haven’t they

140

Reposted by Nathalie Maréchal, PhD

gamera obscura @ohrobin.bsky.social · 01/10/2026

I’m a Pilot. Landing Isn’t as Important as We Think. I’m a Structural Engineer. We’ve Become Too Focused on Whether Buildings Stay Up. I’m a Lifeguard. There’s More to Swimming Than Surviving It. I’m a Historian. The Past is Yesterday’s News.

712824

Nathalie Maréchal, PhD @nathaliemarechal.net · 01/10/2026

I totally get it. This is a bigger problem than any of us can solve individually. Not when we’re up against incredibly rich, powerful companies with all the PR firepower.

010

Reposted by Nathalie Maréchal, PhD

Jos Gallacher 🕊️ @jos.gallacher.be · 01/10/2026

It shouldn’t need saying, but it does.

011

Reposted by Nathalie Maréchal, PhD

Jordan Carlson @jordantcarlson.bsky.social · 28/09/2026

It helps to remember that the entire concept of rationalism was dreamed up by a middle school drop-out who calls himself an autodidact and thinks formal education is for losers.

0315

Nathalie Maréchal, PhD @nathaliemarechal.net · 01/10/2026

That there is, or could be, a “digital mind” of some kind in LLM chatbots

000

Nathalie Maréchal, PhD @nathaliemarechal.net · 01/10/2026

You got the causality backwards there, buddy

000

Nathalie Maréchal, PhD @nathaliemarechal.net · 01/10/2026

💯 - don’t get me started on who gets to claim the title of “scientist” and “AI expert”

020

Nathalie Maréchal, PhD @nathaliemarechal.net · 01/10/2026

One of these days I’m going to bang out a piece about how all the polysemic word games we’ve all been playing for the past however many years (starting with the vague, meaningless “ AI”) gets in the way of productive discussion and debate. Don’t say I didn’t warn you.

3385

Nathalie Maréchal, PhD @nathaliemarechal.net · 01/10/2026

Then there’s the 3D chess logic of “let’s ride the ‘AI Safety’ wave and try to co-opt it to mean the things we’ve always cared about” — all that’s done is confuse the meaning of terms like “safety,” “risk” and “catastrophic” because different people mean wildly different things by those words.

1293

Reposted by Nathalie Maréchal, PhD

Joe Bak-Coleman @jbakcoleman.bsky.social · 01/10/2026

Don’t let the doomer/boomer cults become a partisan sticking point.

171

Nathalie Maréchal, PhD @nathaliemarechal.net · 01/10/2026

Another set seem to be flirting with getting in on the EA funding grift. All I can say is, if that’s your org, you’re going to look really bad a few years from now. Possibly sooner.

2374

Nathalie Maréchal, PhD @nathaliemarechal.net · 01/10/2026

Some progressives (ahem, Bernie) appear to have fully been taken in by the cult of AI. Others seem confused about the difference between the “computer god —> paper clips” scenario vs “shitty agentic AI fucks things up bc the company was careless” - which is what has been happening lately.

19218

Nathalie Maréchal, PhD @nathaliemarechal.net · 01/10/2026

Careless companies doing careless things with half-baked tech products, resulting in real world harm - absolutely a thing. That is not “x-risk.” In the cult of AI, “x-risk” = a more-powerful-than-people synthetic entity autonomously decides to do something that wipes out humanity.

17315

Nathalie Maréchal, PhD @nathaliemarechal.net · 01/10/2026

I absolutely know who needs to hear this, though I won’t name names. Progressive orgs need to stop pretending that “AGI” and “x-risk” are ideas to take seriously. No one is building a computer god that could go rogue and turn us into paper clips.

721172

Nathalie Maréchal, PhD @nathaliemarechal.net · 01/10/2026

This is going to be my new example in media trainings of what NOT to say around a journalist.

010

Reposted by Nathalie Maréchal, PhD

Elizabeth Spiers @espiers.bsky.social · 01/10/2026

Now seems like a good time to say that the more I look at what doge did with USAID the more it looks like a straightforward eugenics project and Elon’s protests that he didnt know anyone might die seem doubly evil

501338417

Reposted by Nathalie Maréchal, PhD

Khashoggi's Ghost @urocklive1.bsky.social · 30/09/2026

It's still the Department of Defense, still the Gulf of Mexico, still Lake Ontario, still Twitter, and still AI. Fuck this guy who thinks everyone else has to follow his weird-ass made up rules about what we call things. Don't call them that now, and officially change them all back when he's gone.

3533260

Nathalie Maréchal, PhD @nathaliemarechal.net · 30/09/2026

Can we all agree that anyone who joins this stupid “SI” bandwagon is a fash loser ? Authoritarianism 101: the regime pressures people to adopt its preferred language as a loyalty test, AND as another step on the slippery slope to subservience. See: Gulf of Mexico. Lake Ontario. Need I go on?

0226