iiinova — where your digital self keeps evolving

Joe Bak-Coleman @jbakcoleman.bsky.social · 13h

It’s gonna be real fun explaining to students the economic bubble propped up on lies that a technology might destroy the world invented by a loser and popularized in his Harry Potter fanfic.

2232

Reposted by Joe Bak-Coleman

Gautam Kamath @gautamkamath.com · 16/09/2026

Nihar Shah did a heroic experiment for TMLR: he spent 20-25 hours over two weeks interviewing authors of seemingly low-quality submissions about their own papers. He confirmed what we all suspected: people submitting these papers have *no idea* what is going on in them.

5293117

Joe Bak-Coleman @jbakcoleman.bsky.social · 16h

Great to see rate limiting at arXiv. A much needed friction.

0110

Joe Bak-Coleman @jbakcoleman.bsky.social · 04/10/2026

Because it keeps popping up in AI articles, “Cure for cancer” is a pretty good indicator that someone knows fuck all about biology. At minimum both should be plural!

24612

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

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

3141

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

Just reviewed a paper with many typos in it and ample use of em-dash. I realized I can use em-dashes again---I love them---if I just dont check my work for typos.

2210

Reposted by Joe Bak-Coleman

Kevin O’Neill @kevingoneill.github.io · 02/10/2026

there is a reason that when we started a grad student methods meeting at Duke, one of our first posts was this fantastic tutorial on GitHub Pages by @khoudary.bsky.social it’s easier than you think! dibsmethodsmeetings.github.io/academic-web...

dibsmethodsmeetings.github.io

How To Build a Free Academic Website

with Jekyll and GitHub pages!

2152

Reposted by Joe Bak-Coleman

Will Lowe @conjugateprior.org · 01/10/2026

If I recall the polisci literature this one of the more plausible theories of lobbying generally: lobbyists aren't there to change minds and generally don't. They're there to provide arguments and data to motivate doing you wanted to do to start with.

061

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

This is, more or less, how industry influence works in science. Just find the scientists who earnestly align with your interests and let them do their own thing. They’re not “on the take” but it just so happens what they wanted to do anyway is easy to fund.

2245

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

Just to put a fine point on it, we add like a teaspoon of baking soda to the fishy euthanasia water so it forms a ph buffer and doesn’t burn their gills. After they’re unconscious, they go into a bag of the stuff (so they won’t wake up) and into the freezer (to ensure death and store for disposal).

1113

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

A grim feature of the death penalty in the US is that it *cannot* be performed by medical professionals, because they're obliged to do no harm. Not only is it randos, but drugs used in MAID aren't generally available because it's a bad look for pharma companies for a *very* niche market.

3219

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

This roller coaster of a year marks fifteen(!!) married the coolest, kindest person on earth.

2470

Reposted by Joe Bak-Coleman

Kieran Healy @kjhealy.co · 01/10/2026

RIP to everyone’s favorite p<0.05 factory

120837

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

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

171

Reposted by Joe Bak-Coleman

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.

721272

Reposted by Joe Bak-Coleman

Tressie McMillan Cottom @tressiemcphd.bsky.social · 29/09/2026

Believe it or not, a lot of people do not have to-do lists. And the the lists many of them have are not structured for an ai agent to be helpful. This isn’t groundbreaking. If you have ever watched the evening news or read even a single story about people who aren’t like you, it is pretty basic

3187299

Reposted by Joe Bak-Coleman

Tressie McMillan Cottom @tressiemcphd.bsky.social · 29/09/2026

So you know previous waves of technological fabulism the tech companies threw a little loose change at social behavioral research and user interface experience. whats interesting about AI interesting is that they aren’t even pretending to care about how people use the technology.

1159684

Joe Bak-Coleman @jbakcoleman.bsky.social · 30/09/2026

“I have a super intelligent model, she lives in a data center in Canada… you probably wouldn’t know her”

0121

Joe Bak-Coleman @jbakcoleman.bsky.social · 29/09/2026

As scientists we don’t often consider or lean on how our humanity gives us skin in the game, and what value that provides to the public. We can face professional consequences for cheating or getting it wrong; it’s that skin in the game that breeds trust. LLM Prof Theseus doesn’t have that.

1145

Joe Bak-Coleman @jbakcoleman.bsky.social · 29/09/2026

"Rather, it comes from interactions within a community of people trained to interrogate evidence and answer for the conclusions they draw from it. And it is time to formalize what this means, and which sorts of human judgment cannot (or should not) be offloaded to AI."

091

Joe Bak-Coleman @jbakcoleman.bsky.social · 29/09/2026

Update on this, I'm on day 2 of my computer basically being out of commission for other science use while it runs these simulations designed for a server inside of a docker container. It's that or rewrite the code for Apple Silicon, updating oh-so-many python packages and not know what changed.

070

Joe Bak-Coleman @jbakcoleman.bsky.social · 29/09/2026

My wife and I had almost exactly this conversation.

1222

Joe Bak-Coleman @jbakcoleman.bsky.social · 28/09/2026

Calling Bullshitter ^I'd actually read this one if @carlbergstrom.com and @jevinwest.bsky.social would write it.

070

Reposted by Joe Bak-Coleman

David Ho @davidho.bsky.social · 27/09/2026

The techbros don't want you to see this trailer. Pass it on.

3249672750

Joe Bak-Coleman @jbakcoleman.bsky.social · 28/09/2026

A rare example of using Betteridge's law of headlines well when reporting on AI. "really... on Its own?" 🤌

0132

Reposted by Joe Bak-Coleman

Carl T. Bergstrom @carlbergstrom.com · 27/09/2026

Ignore for a moment the question about whether Anthropic trains on user sessions. I am I right in reading this as saying that Anthropic's "never-before-discovered" reverse transcriptase was *patented* three years ago?

Dr. Mestre first encountered the enzymes in jumbo phages in 2022 while searching for new reverse transcriptases. He worked briefly as a consultant for ReNegade Therapeutics, which filed a patent in 2023 on a number of reverse transcriptases, including the ones Dr. Mestre found. He is listed as a co-inventor on the patent filing.

10380112

Joe Bak-Coleman @jbakcoleman.bsky.social · 27/09/2026

Had my first email arising from what seems to be an LLM-guided spite-check of my published code. Anyone else experienced this yet? I'm pulling the repo and verifying the possible error, will happily correct if need be but this seems like a great way to DDoS scientists. #metascience #AcademicSky

10556

Reposted by Joe Bak-Coleman

Muireall Prase @muireall.space · 27/09/2026

I think this paragraph on funding entanglement is an understatement. Open Phil made a $30 million grant to OpenAI while Christiano and Dario Amodei were employed by OpenAI, advisors to Open Phil, and living with Karnofsky, at the time engaged to Daniela. web.archive.org/web/20250106...

forecasting researcher Katja Grace co-founded AI Impacts with her then-boyfriend Paul Christiano, a former OpenAI alignment lead, and later dated Alexander. Christiano is connected by friendship and past romance to Holden Karnofsky, the philanthropist whose Open Philanthropy steered hundreds of millions into AI safety, and Karnofsky married Daniela Amodei, co-founder and president of Anthropic. No wrongdoing necessarily follows from any of that, but it does raise significant questions around how AI safety concerns are in fact being handled or promoted with so many industry movers and shakers are entangled to this
degree.

2336

Reposted by Joe Bak-Coleman

Bruno Rodrigues @brodriguesco.bsky.social · 27/09/2026

T 0.55.3 is out 🚀 The latest release of the T programming language brings new data verbs, smarter tooling, and dozens of fixes that make reproducible polyglot data science pipelines more robust! tstats-project.org

tstats-project.org

T Programming Language - index

0111

Reposted by Joe Bak-Coleman

Os Keyes @okeyes.bsky.social · 15/09/2026

Since we’re doing another round of “AI will destroy humanity” doom-hype, it’s a good time to announce a new paper myself and my postdoc advisor just had accepted by BioSocieties on just this. TL;DR: almost all the “AI can make bioweapons” literature comes ultimately from the Effective Altruists.

ironholds.org

17746271

Joe Bak-Coleman @jbakcoleman.bsky.social · 26/09/2026

The journalist covering this preprint is on a fellowship funded by coefficient giving and Longview philanthropy, among others. The lead author of the study Valen Tagliablue is listed as a Digitial Sentience Consortium and Oxford FIG. Who funds those?

32115

Joe Bak-Coleman @jbakcoleman.bsky.social · 26/09/2026

Deeply troubled of course by a lot in this article but it’s wild seeing the character from Cade Metz story, Cameron Berg, become a “computer scientist” here.

312031

Reposted by Joe Bak-Coleman

Max Kozlov @maxkozlov.bsky.social · 25/09/2026

SCOOP: Top researchers at Columbia, Stanford, and others thought they were elected to elite scientific societies alongside Nobel laureates. The catch? The organizations are a sham. I spent the last month diving deep into a massive academic scam spanning continents, for @nature.com

nature.com

Exclusive: Sham scientific societies are misleading star researchers

A Nature investigation reveals a network of organizations that have been building their ranks through deceptive practices.

10309186

Reposted by Joe Bak-Coleman

Naomi Oreskes @naomioreskes.bsky.social · 24/09/2026

Do you study disinformation? Now there is a peer-reviewed journal, dedicated to understanding it. journals.sagepub.com/home/DIN

journals.sagepub.com

Sage Journals: Discover world-class research

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

26924

Reposted by Joe Bak-Coleman

Carl T. Bergstrom @carlbergstrom.com · 23/09/2026

Thinking back to the good old days when publishing a paper every three days required some human to be insecure enough to write up two ideas that didn't deserve to see the light of day every single week.

212621

Reposted by Joe Bak-Coleman

Shannon Vallor @shannonvallor.bsky.social · 24/09/2026

It is a profound indictment of most media coverage of Big Tech that your best bet for getting an informed understanding of AI and its dangers is following a decently curated list of Bluesky shitposters

930686

Joe Bak-Coleman @jbakcoleman.bsky.social · 24/09/2026

Goes rogue in the sense that someone's off-leash dog shits on their neighbors yard.

28312

Reposted by Joe Bak-Coleman

Ersatz Doctor @ersatzdoctor.bsky.social · 23/09/2026

Just taught about Berkeley and got to do my favorite schtick where I ask the students to tell me what they mean by “matter” and then I just keep replying “Stuff *outside* my mind? I have no idea what that could possibly mean.”

2194

Joe Bak-Coleman @jbakcoleman.bsky.social · 24/09/2026

George Berkeley was such a hater and would have loved social media. Writing an entire book hating on Newtons calculus, but subtweeting him by calling it “The Analyst: A Discourse Addressed to an Infidel Mathematician” en.wikipedia.org/wiki/The_Ana...

3142

Joe Bak-Coleman @jbakcoleman.bsky.social · 23/09/2026

A sincere thanks to R1 for making their comments few enough and simple enough on this final round of accepted with minor revisions for me to go on a ride and catch the sunset.

1210

Joe Bak-Coleman @jbakcoleman.bsky.social · 23/09/2026

A good read from @zey.bsky.social. Related to @arxiv.bsky.social 's oral exams: i) chalk talks are going to be really interesting during interviews. ii) Peer reviewing, I've also noticed folks winding up further out over their skis with stats.

1154