Sign in

Mel Andrews

@bayesianboy.bsky.social
13K followers 5.6K following 1.3K posts

I’m not like the other Bayesians. I’m different. Thinks about philosophy of science, AI ethics, machine learning, models, & metascience. postdoc @ Princeton.

PostsRepliesMedia
Mel Andrews @bayesianboy.bsky.social · 18/09/2026
First submit of the term, let me know if you want to talk data models (or read a draft!).
0171
Mel Andrews @bayesianboy.bsky.social · 17/09/2026
“Who wants to see the new Love Island?” Quiet babe my favorite show is on:
Discovering Hidden Relationships of Alloy
Properties with Manifold Learning
Mohammad Abu-Mualla', Ellis Crabtree?, Fredrick Michael, Yayue Pan', Jida Huang'*
'University of Illinois Chicago, 842 W. Taylor Street, Chicago, 60607,
Illinois, United States.
NASA Marshall Space Flight Center, Martin Rd SW, Huntsville,
35808, Alabama, United States.
Q=
e.
*Corresponding author(s). E-mail(s): jida@uic.edu;
Contributing authors: mabumu3@uic.edu; ecrabtr2@jh.edu
fredrick.n.michael@nasa.gov; yayuepan@uic.edu*;
090
Mel Andrews @bayesianboy.bsky.social · 15/09/2026
……In perfectly unrelated news, have an advert…….
180
Mel Andrews @bayesianboy.bsky.social · 11/09/2026
You can now read this invaluable book on methods in philosophy of science open-access. Find me in Chapter 16!
0364
Mel Andrews @bayesianboy.bsky.social · 01/09/2026
0342
Mel Andrews @bayesianboy.bsky.social · 26/08/2026
I, too, have been experiencing a sense of alienation from my discipline. Not because I don’t believe that philosophy strives for truth and objectivity, but because believing that the sort of pseudo-reasoning LLMs engage in can achieve these epistemic goods requires having a bad epistemology.
Thread from Megan Fritts reading “For me, the biggest source of grief arising from the AI crisis hasn’t even been in the realm of teaching. It has been the realization that a non-trivial number of my disciplinary peers see philosophy in a way that is wholly different from—even antithetical to—the way I see it.”
1669
Mel Andrews @bayesianboy.bsky.social · 21/08/2026
How disappointing. bsky.app/profile/baye...
081
Mel Andrews @bayesianboy.bsky.social · 21/08/2026
Substack: do you want to use an AI tool for writing to check your writings about not writing with AI for AI-written writing?
1233
Mel Andrews @bayesianboy.bsky.social · 20/08/2026
These are the top google image search results for android epistemology
1121
Mel Andrews @bayesianboy.bsky.social · 20/08/2026
Does anyone happen to have a text-searchable PDF of Android Epistemology (1995) lying around?
3176
Mel Andrews @bayesianboy.bsky.social · 14/08/2026
I have learned in one and the same breath that Cambridge’s Jason Arday is dead and that “Blacks would disappear in a meritocracy” Nathan Cofnas was at the head of the witch hunt against him.
5412
Mel Andrews @bayesianboy.bsky.social · 11/08/2026
Look at the kooked out OCR github has implemented. Look how they’ve yassified my {LaTeX}.
380
Mel Andrews @bayesianboy.bsky.social · 10/08/2026
According to consumer reports, users love the way Dr. Andrews can conduct high-impact research, innovate in the classroom, and look mostly not pissed off during Q&As.
1201
Mel Andrews @bayesianboy.bsky.social · 10/08/2026
Honored to have @kathleencreel.bsky.social @vlastasikimic.bsky.social @eduede.bsky.social @sfazelpour.bsky.social @lastpositivist.bsky.social @angelinawang.bsky.social @mjcrockett.bsky.social @niallroe.bsky.social on board!
051
Mel Andrews @bayesianboy.bsky.social · 10/08/2026
Excited to be holding a symposium on AI/ML in science evaluation at PSA 2026. Who here is going to San Diego? @philsci.bsky.social
4224
Mel Andrews @bayesianboy.bsky.social · 05/08/2026
Among the better words I have read in recent times.
0220
Mel Andrews @bayesianboy.bsky.social · 04/08/2026
You ever really enjoy an apéritif and then read the fine-print…?
2222
Mel Andrews @bayesianboy.bsky.social · 31/07/2026
Growing up, I always thought of black squirrels as “Harvard Squirrels” owing to my experience of their natural distribution. Little did I know, the melanistic eastern gray squirrel is endemic to ivy league colleges generically. Only the Harvard Squirrels at UPenn, however, come when called.
harvard squirrel
5320
Mel Andrews @bayesianboy.bsky.social · 10/07/2026
The cats do not understand why they are not permitted to lick the mouse.
1180
Mel Andrews @bayesianboy.bsky.social · 10/07/2026
The mouse now has an incurable exercise addiction.
3130
Mel Andrews @bayesianboy.bsky.social · 10/07/2026
In the early stages of drafting a manuscript, it can be helpful to draw a little mascot (mousecot?) to cheer you on.
20845
Mel Andrews @bayesianboy.bsky.social · 09/07/2026
is this anything
4260
Mel Andrews @bayesianboy.bsky.social · 02/07/2026
I do wonder for how much longer I will have to explain this at least once per day.
713646
Mel Andrews @bayesianboy.bsky.social · 27/06/2026
home at last
3170
Mel Andrews @bayesianboy.bsky.social · 17/06/2026
The irony of using an LLM-powered bot with a female-coded screen name to respond to an observation about gender bias in LLMs is not lost on me.
1301
Mel Andrews @bayesianboy.bsky.social · 17/06/2026
Interestingly, the models themselves offer pseudo-mathematical explanations for this effect, claiming that women authors are simply “less salient.”
181244188
Mel Andrews @bayesianboy.bsky.social · 11/06/2026
Whether predictive systems can be ethically—or even functionally—used in socially sensitive contexts depends on their epistemic credentials. I am delighted to see my work inform policy recommendations in a new report from Amnesty International. www.amnesty.org/en/latest/ne...
Cover of Amnesty International Report – AUTOMATING SUSPICION
RISK PROFILING AS A SMOKESCREEN FOR STRUCTURAL
DISCRIMINATION AND INEQUALITY
1286
Mel Andrews @bayesianboy.bsky.social · 11/06/2026
Gmail very insistent these days that all my emails should be translated to Polish. Any way to turn this feature off (or change language)?
4121
Mel Andrews @bayesianboy.bsky.social · 05/06/2026
what do you think it would take for me to break into the AI Generated Artificial Fake Bean Plastic Vegetable Party market
080
Mel Andrews @bayesianboy.bsky.social · 05/06/2026
170
Mel Andrews @bayesianboy.bsky.social · 05/06/2026
guys I have made an important discovery
false beans. beans. made of plastic.
5150
Mel Andrews @bayesianboy.bsky.social · 15/05/2026
So no one catches my vibe when I say DimeNet filters (Directional Message Passing Neural Network filters for distance-wise and angular atomic positioning in molecular structure learning) look like finch chick gapes?
1131
Mel Andrews @bayesianboy.bsky.social · 13/05/2026
I think often about the United States as an entity willing to commit war crimes against its own citizens on its own soil. Today marks 41 years since the MOVE bombing. It took place here, in West Philadelphia. And America remains a nation constitutionally at war with itself.
2346
Mel Andrews @bayesianboy.bsky.social · 30/04/2026
The fantasy of automating science is premised on the idea that science can be eliminated—indeed, that it strives for its own elimination—in favor of engineering. That the work of science is an avoidable inconvenience. Thus the ‘mistake’ in what science is taken to be is judged as robbing humans of
A final angle on why its fraud might be that it misportrays what science is more broadly. If you think the reasons to do science are fundamentally human-that as scientists we are concerned with producing understanding for ourselves just as much as we're concerned about improving things in the world-then you could argue that for science to be meaningful we have to be the ones coming up with the ideas and shaping them as we go. From this perspective, automated science isn't inherently wrong, it's just missing the point. Al for science arguments often completely overlook the "people production" role of science. In the extreme, they envision Al finding solutions for lots of real world problems and intervening to control outcomes in the world without us understanding how any of it works. In reality, the personal side-including the search for personal fulfillment through science-is a big part of why smart people who could potentially make a lot more money in applied roles end up choosing research careers. And it's a big part of how we evaluate scientists. How many of your Ph.D. students have gone on to competitive research positions? What does the trajectory of topics you've worked on say about your research taste?
1216
Mel Andrews @bayesianboy.bsky.social · 23/04/2026
In recognition of my love for the no free lunch theorems PennDOT bestowed on me these complimentary glam plates
180
Mel Andrews @bayesianboy.bsky.social · 23/04/2026
This is an important read if you want to understand Palantir’s modus operandi and how silicon valley has accomplished near total corporate capture of government.
23911
Mel Andrews @bayesianboy.bsky.social · 12/04/2026
The answer to the first question is: it remains a game of economic domination and regulatory capture.
2404
Mel Andrews @bayesianboy.bsky.social · 12/04/2026
Why is pro-AI still posturing as anti-AI, and why is the general public still willing to believe the charade?
76912
Mel Andrews @bayesianboy.bsky.social · 01/04/2026
Decades of activism fought to make American cities handicap-accessible by installing curb cuts and tactile pavement. Delivery robots default to stopping in those curb cuts, effectively eliminating access to public walkways for those using mobility aids.
delivery robot parked in curb cut adjacent to zebra stripes
36418
Mel Andrews @bayesianboy.bsky.social · 14/03/2026
…and on American dime, no less…
171
Mel Andrews @bayesianboy.bsky.social · 04/03/2026
yet I settled for a lemonless existence
0110
Mel Andrews @bayesianboy.bsky.social · 04/03/2026
First time in Berkeley since nearly coming here for my PhD. A return to the land of giant lemons.
4420
Mel Andrews @bayesianboy.bsky.social · 27/02/2026
In my imagination people bet on the outcomes of bingo games but I may be wrong about this.
230
Mel Andrews @bayesianboy.bsky.social · 25/02/2026
is this the future we want?
2240
Mel Andrews @bayesianboy.bsky.social · 20/02/2026
Sigh. I love to be cited, but love even more when people read what they are citing. I’ve argued that P1. ML is inductive inference, P2. All inductive inference requires theoretical input, C. ML requires theoretical input for inference. This is all well-rehearsed. Far more interesting than this is a
2211
Mel Andrews @bayesianboy.bsky.social · 18/02/2026
This new writing style was not merely a spectacular innovation; it was a prescient challenge to the status quo of literary orthodoxy.
0252
Mel Andrews @bayesianboy.bsky.social · 18/02/2026
partner sees picture of colugo for the first time: that is a demonic flying book
1150
Mel Andrews @bayesianboy.bsky.social · 17/02/2026
Something about LLM hype culture renders a man immune to the experience of embarrassment. If I couldn’t tell the difference between PhD-level scholarship and grammatical gibberish I simply would not announce that to a global audience.
613920
Mel Andrews @bayesianboy.bsky.social · 11/02/2026
Princeton University has just unveiled its new mascot. I am now fully and unreservedly in support of school spirit.
060
Mel Andrews @bayesianboy.bsky.social · 11/02/2026
they already went far enough painting the squirrels
180