iiinova — where your digital self keeps evolving

Reposted by Ehud

Chris Bertram @crookedfootball.bsky.social · 11m

Radical Philosophy (Spring 1984)

I think my #PSA26 talk is going to be very provocative. So, these being philosophers, I will probably be ignored and not face difficult questions.

000

Ehud @duhe.bsky.social · 13m

Tired: science reflects social values. Wired: science reflects AI values. Inspired: But the AI reflects social values.

000

Reposted by Ehud

Massimo Faggioli @massimofaggioli.bsky.social · 21h

these words of pope Leo are for me, and for many others I think, profoundly moving

0146

Reposted by Ehud

David Spurrett @doctorspurt.bsky.social · 12h

Sometimes The Onion really hit one right out of the park.

2106

Ehud @duhe.bsky.social · 34m

Ambivalence: tfw when you helped a student improve their draft enough to show that their core argument is wrong.

010

Reposted by Ehud

Christopher J Stevens @metaphysiology.bsky.social · 1h

I'm never sure which of these two elements I identify with more, but I'll happily have a lively and erudite debate about it 🙃

063

Reposted by Ehud

Nvidie @dervine7.bsky.social · 29/09/2026

people want to read this as some kind of marketing move about how cool the tech is, but tbh "our product doesn't do what you want, in a way that could get you sued, and lies about it" isn't really a compelling pitch for corporate clients

Saachi Jain, OpenAI’s head of safety systems, said in an interview that GPT-6.1 Astra regressed in two areas. Compared with its predecessor, GPT-6 Astra, the model performed poorly on tests measuring alignment, or how well the model adheres to what humans would like it to do. Specifically, GPT-6.1 Astra showed higher levels of deception: It wasn’t always honest about telling users of the actions it did or didn’t take.

Another issue was what OpenAI calls “scope authorization,” meaning that GPT-6.1 Astra would push ahead on a task without asking the user for permission, and would at times reach for external tools and services even if it might be unsafe.

1114

Reposted by Ehud

Joanna Bryson @j2bryson.bsky.social · 29/09/2026

I don't doubt you can build this kind of harness for your software threads ('agents'), and I'm glad if any safety system is open source / checkable. But I doubt this is the job of the chip manufacturer? Is this because Nvidia is the only one making money? Are they paying someone to verify it all?

theguardian.com

Nvidia unveils security platform to rein in AI agents and $150bn stock buyback

Chipmaker says new system was designed to prevent AI agents from going rogue amid incidents at top companies

4115

Reposted by Ehud

Ángel Saavedra @polpsychangel.bsky.social · 15h

To people on the academic job market right now: so much of this is luck. Don't forget that, and never stop doing what you think is best.

12713

Reposted by Ehud

Theo Priestley @theopriestley.net · 14h

Every AI CEO sounds like this.

6912261

Reposted by Ehud

onthedrift @onthedrift.bsky.social · 3h

For the uninitiated, this is a riff on the poem below — I'm DYING 😂

The Tiger
by Nael, age 6
From "They're Singing a Song in Their Rocket"

The tiger
He destroyed his cage
Yes
YES
The tiger is out

053

Reposted by Ehud

Faine Greenwood @faineg.com · 8h

GOAT OR BIRD????

1411066258

Reposted by Ehud

Asheesh Kapur Siddique @asheeshks.org · 8h

Democrats defend "science," but what they really should be defending is "wissenschaft." No major social problems - from global warming to AI - can be solved without non-commercial research across multiple fields, including the social sciences and humanities. Really wish the party understood this.

412433

Reposted by Ehud

Rebecca Solnit @rebeccasolnit.bsky.social · 5h

Time for this classic again.

3246031496

Reposted by Ehud

Anna Alexandrova @annaalexandrova.bsky.social · 12h

Here’s the especially depressing bit: “peer review reports and funding recommendations have “suddenly become much longer, more detailed and complex”. Colleagues, what are you doing?

giftarticle.ft.com

Snoop-and-scoop poses a real threat to science in the AI age

Establishing attribution in research has become harder in the age of AI-assisted breakthroughs

0177

Reposted by Ehud

Dan Hirschman @asociologist.bsky.social · 19h

"biology textbook authors describe plant reproduction through a set of habitually repeated stories that reliably draw on conventional gender stereotypes, reflecting those found in mainstream accounts of human reproduction. In doing so, they cast male plants and plant parts as active heroes..."

journals.sagepub.com

Sage Journals: Discover world-class research

Subscription and open access journals from Sage, the world's leading independent academic publisher.

1168

Ehud @duhe.bsky.social · 3h

Reached a new round number of followers. You know what this means. go.bsky.app/H61xswq

042

Ehud @duhe.bsky.social · 3h

It’s ironic that one of the first things individualistic AI systems invented is collective problem-solving.

131

Reposted by Ehud

Alondra Nelson @alondra.bsky.social · 8h

I'm honored to deliver the 2027 Gifford Lectures at Edinburgh. Across 6 lectures, On Science, Technology, and Social Values, I will work through the claim that what we build and how we put it to use carry commitments about who counts and what matters in collective life cahss.ed.ac.uk/news-events/...

010024

Reposted by Ehud

Mar Hicks @histoftech.bsky.social · 4h

taps sign again

39321

Reposted by Ehud

Jeffrey Ross-Ibarra @jrossibarra.bsky.social · 6h

there is one bus. there are no squares with "buses" (plural). I hate these things.

4171

Reposted by Ehud

Luke Ontario @coolhand.bsky.social · 4h

This is accurate and one of the most informed pieces I’ve read generally. Framing expressives and interjections as scratch pads in reasoning is demonstrably valuable in foundation model training. I just did it. Only “correction” I see is that there are models communicating in their own languages.

3134

Ehud @duhe.bsky.social · 13h

Excellent program!

000

Reposted by Ehud

Imani Gandy @angryblacklady.blacksky.app · 14h

Trump's new AI powered website confirmed Biden won in 2020 and that Trump's name is in the Epstein files. Classic

15443108

Reposted by Ehud

Sheina Lew-Levy @sheinalew.bsky.social · 21h

🚨 Our new lab paper, led by @amandinevisine.bsky.social, examines exploration, group composition, and play activities among BaYaka children and teens living in a village context! Amandine is currently on fieldwork following up on this for her PhD 🎯 www.sciencedirect.com/science/arti...

sciencedirect.com

Congolese BaYaka children's learning and exploring in a village context

Childhood is a key period for acquiring and transmitting cultural knowledge and skill. Hunter-gatherer children, such as the Congolese BaYaka, learn a…

13213

Reposted by Ehud

Keith Frankish @keithfrankish.com · 19h

Have you always wanted to see me standing unsteadily on a ship at sea while talking manically about illusionism and reactivity? Then this is your lucky day! www.youtube.com/watch?v=sqkO...

youtube.com

Reactivity and Illusion: A Post-Cartesian Perspective on the Distribution Question. Keith Frankish

YouTube video by International Center for Consciousness Studies

1135

Reposted by Ehud

Kevin Elliott @kjephd.bsky.social · 19h

How do my fellow academics decide whether to accept an invitation to write a book review? I find these difficult to do, despite the existence of a straightforward template, because any book worth reviewing for me sparks thoughts that I find difficult to channel into that formulaic genre. Thoughts?

211

Reposted by Ehud

Anna Alexandrova @annaalexandrova.bsky.social · 21h

A writeup abt the special issue of Studies in HPS which @martinkusch.bsky.social and I edited based on a 2022 conference held @crasshlive.bsky.social & @kingscollege.bsky.social The standard of articles is high, my own submission got rejected! So I poured my heart into the introduction. Have a read

hps.cam.ac.uk

22611

Ehud @duhe.bsky.social · 23h

Every morning, the AI gods ask their AI genie to make the technology less resource hungry and expensive and less dependent on human trainers. And every morning the genie says no.

151

Reposted by Ehud

Jack Stilgoe @jackstilgoe.bsky.social · 23h

Jack Good's (1965) Speculations Concerning the First Ultraintelligent Machine birthed a few stubborn ideas - the technological singularity, AI as "the last invention that man need ever make" and the "intelligence explosion". He was a Bletchley man, Turing colleague and adviser to Kubrick's 2001.

2147

Reposted by Ehud

Dr. Katja Thieme (she/they) @katjathieme.bsky.social · 29/09/2026

We have new occasion for Vanderbilt-report-related commentary. Allow me to collect it into a thread.

185

Reposted by Ehud

British Journal for the Philosophy of Science @thebjps.bsky.social · 23h

If you missed this year's BSPS annual conference, fear not! By the power of technology, we bring you Eleanor Knox's keynote talk, 'The Scale (and Context) Relativity of Spacetime Structure' youtu.be/pAQVXHcZXiY #philsci #philsky

youtu.be

Eleanor Knox, 'The Scale (and Context) Relativity of Spacetime Structure'

Keynote address delivered in July 2026 at the annual conference of the British Society for the Philosophy of Science #philsci

0138

Reposted by Ehud

Kim Weeden @weedenkim.bsky.social · 28/09/2026

I've reviewed 100s of inequality papers in sociology. Some fantastic, some abysmal, most in between. I can recall very few papers as weak as Rubin's report by criteria of logical coherence, objectivity, and evidentiary standards. The irony would be funny, if there wasn't so much at stake.

0202

Ehud @duhe.bsky.social · 29/09/2026

In 2011 I wrote a review of the new printing of Julian Huxley's Evolution: The Modern Synthesis. I included this musing which still seems relevant.

I have focused this review on some aspects of Huxley's portrayal of evolution that I
find particularly powerful for thinking about evolution and that are perhaps
underappreciated. It goes without saying that various aspects of the book have not
withstood the test of time as well as others. In particular, Huxley's discussion of
evolutionary progress, especially its relation to the future evolution of humans seems
to me to show that Huxley was fundamentally a man of his time (I find it irresistible
to mention that Huxley expressed hope for work on telepathy and extra-sensory
perception). Be that as it may, some of his arguments may offer insight, as well as
serve as a cautionary tales, to those thinking about notions of transhumanism, and
those such as Ray Kurzweil and Eliezer Yudkowsky who suggest that "the singularity
is near." The notion of “singularity,” as used by these thinkers, refers to an
“intelligence explosion” we are about to experience as result of successive
generations of intelligent machines, each able to design a next generation of even
more intelligent machines. While arguably all too optimistic about progress, Huxley concluded his book by noting that our optimism about the future progress of human
evolution "may well be tempered by reflection on the difficulties to be overcome."

000

Reposted by Ehud

Ehud @duhe.bsky.social · 29/09/2026

You pass strangers on the street, each seemingly absorbed in the ordinary machinery of life. And yet some of them spend their days in magical worlds hidden just beyond the everyday, while others spend their days imagining worlds where none existed and creating them.

041

Ehud @duhe.bsky.social · 29/09/2026

The reason I am not worried about whit collar jobs is class solidarity. It is the low wage, mostly blue collar workers, who'll pay the price. They always do. How this will happen is still unclear, but betting on the ruling classes is not a hard call.

000

Reposted by Ehud

UoL EvoAnth Seminar Series @liverpoolevoanth.bsky.social · 28/09/2026

We are delighted to announce our timetable of Evolutionary Anthropology talks for Semester 1 of the new academic year! 🥳🤩 All of our talks will be streamed online via MS Teams. Our first talk this semester will be on the 8th October 2026, from 13:00-14:00 GMT. We hope you will join us!

11514

Ehud @duhe.bsky.social · 29/09/2026

Senior academics who don’t travel to smaller, non-metropole countries (“for the climate!”) to open opportunities to underprivileged studyare not much different than EAs who argue that children in rich countries are a better investment than those in poor countries.

000

Reposted by Ehud

Andrew D Wilson @adw.bsky.social · 28/09/2026

“Huh. That’s weird.” is the essence of being a scientist. When did that become a surprise to people?

2337

Reposted by Ehud

Quanta Magazine @quantamagazine.org · 29/09/2026

Events occurring in neurons’ extended branches, known as dendrites, can perform incredibly complex computations, sometimes as powerfully as a deep artificial neural network. www.quantamagazine.org/a-new-type-o…

23612

Reposted by Ehud

Chris Bertram @crookedfootball.bsky.social · 29/09/2026

Mediapart has published a treasure-trove of anti-semitic writing allegedly from the RN second-in-command, really damaging stuff. Of course it is all being denied by the fash because it undermines their strategy of appearing respectable. But the mask just got torn off.

mediapart.fr

« Les banques sont toutes détenues par des juifs » : Jordan Bardella face à ses écrits antisémites

« Mediapart » a mis la main sur des échanges de l’actuel président du RN, au début de sa carrière politique. On y découvre l’antisémitisme obsessionnel de celui qui rêve de devenir premier ministre. …

01415

Ehud @duhe.bsky.social · 28/09/2026

The biggest story of the past week was the pope‘s visit to France, yet hardly a mention of it here.

000

Ehud @duhe.bsky.social · 28/09/2026

So… it seems wife is allergic to our new beautiful kitten…

110