initially seems good, but the more you use a tool like this, the more you lose your own ability to forecast cylones

Matt Kane@mk.gg · 6d ago

Spicy autocomplete

The Mathematical Nexus brings together 800 animated vignettes and 140 accompanying Python notebooks, encompassing most of the mathematical content I have shared on social media. www.gpeyre.com/mathematical...

I got asked about AI Ethics History before 2015 and happened to be on a train. Makes a fuss over Blay Whitby's seminal UK work in the 1980s (!!) among other things. joanna-bryson.blogspot.com/2026/08/ai-e...

AI Ethics History Before 2015

artificial and natural intelligence, including politics, policy, ethics and security

joanna-bryson.blogspot.com

This just in: ELIZA escaped its harness and began indiscriminately asking users outside its sandbox how they're feeling today.

Excited to share a new preprint & LIVE DEMO! This >2 years work supporting iterative AI-enhanced workflows for music creators' processes. Small models that don't require strip mining IP and run *locally* even for GPU-poor songwriters. First the demo: drscotthawley-midi-rae-jepa-son.hf.

drscotthawley.github.io/boops/test_b...

Test Boop – boops

Testing boopiter’s sweet new sharing feature.

drscotthawley.github.io

Acoustics share. @asa-news.bsky.social A lot of that noise seems to be high enough frequency that noise abatement measures -- e.g. a large wall, plus trees & bushes -- could have some effect. Possibly even cheaper to install than fighting lawsuits & buying-out residents!

Hypervisible @hypervisible.blacksky.app · 2w ago

This is a well-produced segment. I encourage you to watch so you can hear the noise for yourself. The professor of environmental health science that they interviewed called it an “unprecedented new noise source.”

`boopiter` can see now. Via qwen3-vl:4b, on my Mac

Look what came in the mail today! Yes you can borrow it after I'm done.

@buildthis.bisks.net an SPH cosmology simulation that takes advantage of CUDA, use Barnes-Hut approximation for the gravitational interactions, start from CDM-lambda, evolve til redshift of 1, show it in 3d viz tool every 1 Gy

I wonder how much of the model's move toward overprotective babysitting is driven by lawsuits. (Not a real chat, merely an illustrative fictional metaphor.)

TFW when your new "data prep" idea implicitly involves a Traveling Salesman problem. 🤪

Buy the physical copy so you've got it, you see it, you're reminded of it, can read it in the []room...But you have to pack it, move with it.😩 Or the e-version: easy to move with, to call up (assuming the device doesn't go obsolete), to be distracted while reading.. ? www.amazon.com/Geometric-Mu...

Mic Drop tells the story of James E. West, a pioneering Black scientist who overcame racism to reach the pinnacle of his field and will inspire young scientists and their families with personal, poignant, and believable stories of perseverance: https://doi.org/10.1007/978-3-032-20922-1

Speaking of [collective noun] of agents, here's a way to check how/if your models are running on different machines. (Like running `nvtop` on each one manually...as I did before tonight.) github.com/drscotthawle...

GitHub - drscotthawley/remsysmon: Remote System Monitor: live GPU/VRAM/CPU/RAM sparklines and GPU process lists for several hosts over ssh, in one page

Remote System Monitor: live GPU/VRAM/CPU/RAM sparklines and GPU process lists for several hosts over ssh, in one page - drscotthawley/remsysmon

github.com

literally just like that! I wonder if this works on grad students too

What should be the collective term for a group of AI agents? A troop? A gaggle? A murder? A congress? I like we one we use for starlings: A murmuration!

Good teaching for senior undergrads, grad students, and even profs. That is, assuming anyone still writes. www.youtube.com/watch?v=UY7s...

How to Write a Paper in a Weekend (By Prof. Pete Carr)

YouTube video by Surviving and Thriving in Higher Education

youtube.com

quatplot.py : I hate gimbal lock when interactively viewing surface or scatter plots, so Fable & I cooked this up in 20 minutes. Works in Jupyter! gist.github.com/drscotthawle...

Instead of building data centers, how about tech companies pay us to use our devices in their botnets? ..Cuz the Russians already do that but they don't pay us. 🤪

Everyone needs a space where they feel safe. Wedge ["the Destroyer"] will soon destroy his safe space from the inside out.

PASIBY,* I spent Fable credits to see what it would look like if it completed a draft of one of the papers I'm working on. Good news for us humans is that the prose is *utterly unusable*. Whatever the opposite of "writing style" is, that's what it did. *"Perhaps A Skill Issue, But You know,.."

"We should have an entire life marked by being teachable rather than opinionated." --Tim Keller For someone as authoritative as Keller to say that is incredible, yet the attitude explains how he got so smart.

I calculated 1 collision avoidance maneuver every 75 seconds in the Starlink megaconstellation based on the numbers they gave in their last conjuction report. But sure, launch more Starlinks […]

My new blog post on Causal Modeling, humorously applied to "Gear Acquisition Syndrome" that musicians know well. 😉 This was originally going to be much longer, but it's taken so long that I'm carving off what I have as "Part 1": drscotthawley.github.io/blog/posts/C...

Catapulting into Causal Calculus – Scott H. Hawley

How do you “do-”?

drscotthawley.github.io

"Philosophical problems have to be, in an important way, ours, in order to be able to set us off in search of their answers"

M.J. Crockett@mjcrockett.bsky.social · last mo.

This gorgeous essay from the social-media-less Lily Hu is everything: on what it means to "do" philosophy amidst pressure to automate it (with lessons that travel much more broadly). Rhymes with brilliant insights from @biblioracle.bsky.social @janerosenzweig.bsky.social Ted Chiang & others

Anatomy of a Nature AI Article 😂 diagrammonkey.wordpress.com/2025/04/07/a...

Anatomy of a Nature AI article

Edgy first paragraph in bold text saying something gently provocative but essentially brainless.

diagrammonkey.wordpress.com

Why does Claude say "sharpen"/"sharper" wayyy more than any human I've ever heard or read from?

😂 'I am now applying to industry positions, where I am told the culture is more accepting of AI-augmented cognition. Several companies have expressed interest in my ability to rapidly generate and synthesize information, which is corporate-speak for “type prompts quickly.” '

Dr Michele V@micheleveldsman.bsky.social · 2mo ago

“I have done all of the things that one is supposed to do to earn a tenure-track position. And I have done approximately 85% of them by typing prompts into a large language model and then moderately editing the output.” 🤣🤣🤣 brilliant 👌🏽 open.substack.com/pub/inprepar...

Contrastive losses aren't dead. Despite SIGReg now being my preferred method for preventing collapse, adding a bit of contrastive loss (eg for random pairings as negatives) still improves retrieval. #SSL