Cameron Patrick

camping gear ready to pack into a backpack

brb off for extended grass-touching and mountain-gazing

Dubai real estate prices: definitely the most important thing to be worrying about in the middle east right now

Cover image of the 5 March 2026 issue of Science Magazine. Koalas (Phascolarctos cinereus), such as this mother and 7-month-old joey from Queensland, Australia, embody a genetic paradox. Populations rich in diversity are declining, whereas those with little variation are expanding and rapidly reshuffling their genomes. These findings reveal that diversity alone does not determine resilience. Instead, a population’s fate depends on several evolutionary processes unfolding across generations.

Genome sequencing in 418 koalas from 27 populations across Australia shows that, though they still have low diversity due to past decline, there are clear signs genetic recovery is underway.
Escaping bottlenecks: The demographic path to genetic recovery in koalas www.science.org/doi/10.1126/...
🐨🧬😊🧪

new tag yourself just dropped

I cannot let any mention of Graeber pass without observing that he and his politics were soundly defeated by the logistics of… legally owning and operating a car, so instead they destroyed it

neuroplausible: I Hate Matlab: How an IDE, a Language, and a Mentality Harm This blog post is inspired by a few Matlab-related tweets of mine, which turned into days-long discussions with fellow science and non-science tweeps. Those tweets of mine in turn are motivated by two...

Every time I am forced to use Matlab I can only conclude it is a language designed by and for sociopaths

Shout out to this @olivia.science blog post that I sent out on a neuro listserv back in grad school, that a prof there is apparently still mad about

neuroplausible.com/matlab

I remember when you had to choose between internet and telephone, rather than the internet being a thing you do on your telephone

A digital CAPTCHA verification window titled "Select all squares with PIPES" against a plain white background. The window contains a 3×3 grid of numbered squares, mixing literal hardware, smoking pipes, and programming syntax.

These captchas just keep getting harder #rstats

if they're not actual lawyers, then surely they shouldn't be offering legal advice (for a fee, even!)

Slay the Spire 2's concurrent Steam players just hit 179,456, the highest ever for any roguelike.

When StS1 first launched back in 2017, it had 193 concurrent players.

That's a 92,982% increase, meaning StS3 is on track to hit 166,861,777 concurrent players by 2035. 💪

Welcome back *checks notes* ANZAC cove?

The trend shift in adults per family household is so insane.

its even more insane that every important policymaker and economist thought we were slowly soaking up a glut of homes even into the 2010s.

And it’s even more insane that there are still supply-crisis skeptics today.

pandoc 1, claude 0

I'm a big fan of Bertrand Russell's 1932 essay "In Praise of Idleness", which begins by categorising jobs into "moving objects around" vs not, and notes that the "not" group is more pleasant and better paid. We should be aspiring to these jobs, not calling them bullshit!

People say BlueSky is an echo chamber but we have road cyclists, gravel cyclists, mountain bikers, single speeders, fixie riders, cargo bike riders, e-bike riders, hybrid riders, recumbent riders, bikepackers, randonneurs, and I'm sure we even have a few unicyclists among us

who needs a control group when you have AI!

Tbf last night the 7 year was telling me about "cat memes" (business cat, science cat) so maybe we will be culturally relevant again soon

agreed

fucking FINALLY

Then I save-scummed and took the other,

Two fossils diverged in a moonlike cave,
And sorry I could not take both
And be one traveler, long I gave
And looked down one as far as I could
To where it revived in its growth;

Possibly the worst is that it contains a lot of the right jargon and looks kinda superficially similar to good analysis plan but as soon as you start paying attention to the details it's full wtf

I respect this pokesplaining and now feel like a bad millennial that I've never played or watched ANY kind of pokemon. (Not quite sure how that happened tbh, I should really have been prime demographic back in the 90s)

if you're gonna write a paper claiming "LLMs can generate a statistical analysis plan for a clinical trial", and then provide THIS as an example of good output... big yikes

argh

i've seen worse

nevertheless

argh

well I certainly hope you can, because in practice aren't all models misspecified?!

but yes that's a good point, it's often not really discussed and I may have been making incorrect assumptions also...

I was thinking in terms of applying models that assume normality to situations where that assumption doesn't hold - isn't that the same in Bayes or frequentist frameworks, relies on the CLT?

Drop Site News sounds a lot like Drop Table Students

now I'm contemplating inflicting maximum psychic damage by teaching everything in terms of linear models (because It's Just A Linear Model) but interpreting OLS purely as the Bayesian estimator with Jeffreys priors

interesting, why would you remove teaching the CLT in a Bayesian stats course?