testingdeepwaters.bsky.social
@testingdeepwaters.bsky.social
It's just a test.
You get one more shot to show me you're not a troll.

When I said: "I am TIRED of people just brushing them off like ignoring them will make them go away." I was clearly talking about "the "politically extreme" and "populist" assholes". Earlier, I called them "elitist "cranks"".
October 8, 2025 at 7:17 PM
I might stand corrected on this. Thankfully, it's merely an example given to support a conceptual argument by analogy. However, it does provide some indication of how far misunderstandings of biological explanation have spread!
October 1, 2025 at 10:07 PM
If we're "just machines", then we don't have psychologies, and we only just react to stimuli as machines do. But we don't. And strictly-speaking, machines are artificial, while workers aren't. We can draw analogies, but most elitists don't care about the limits of those when talking about us.
August 27, 2025 at 8:31 PM
You first gotta understand what people are talking about. Most scientists and popularizers won't help: they draw analogies but then appropriate ordinary language words and break the empirical and conceptual limits of those analogies. I recommend this book: search.catalog.loc.gov/instances/17...
August 13, 2025 at 12:26 PM
For anyone still here and interested, have a paper with a REALLY good visual about one form of "Goodhart's law". arxiv.org/pdf/2310.09144
August 6, 2025 at 12:29 PM
People can't BE brains, either, since, for example, personhood as a concept applies to whole organisms.

We're lucky there's not just one book on it, but many more from the same authors, and still more from other authors. None of that stops crappy notions spreading in niche and popular science.
July 31, 2025 at 1:15 PM
I found a typo, so your entire blog must be wiped from existence.

Jokes aside, what I screenshot understates the problem! We're often interested in "causality", but what's an is "causality", huh?! (Muphry's law.) Conceptual issues cloud matters even before quantitative analysis can begin!
July 28, 2025 at 3:43 PM
Hahahahahahaha! Pole-vaulting the gun so badly! I expect a mental decathlon from this dupe!
July 24, 2025 at 1:31 PM
I felt like I was trying to help him. Did I not convey that enough?

Whatever the case, I expect a crazed political opportunist to smell blood and write a one or more huge blog posts about how he's not a real scientist and (their main conclusion) how whatever general positions he supports are awful.
July 22, 2025 at 9:50 AM
Now for the paper, DiFrisco & Gawne 2025. So far my head's getting sore from nodding in agreement, but I think something's wrong with their talk about colloquial "agency". Why would it presuppose any neurobiological substrate at all? It's just a suite of powers that a body has.
July 19, 2025 at 6:33 PM
Said untrained eyes might miss this graph.
July 18, 2025 at 9:40 AM
Not a fan of some of the graphs they DO provide. Those look like confidence intervals of means. Why not also provide the sample SD? Untrained eyes will see unequal arousal with no overlap.
July 18, 2025 at 9:38 AM
Pretty much. You can easily find signatures of that in Galton's work.
July 17, 2025 at 9:30 PM
Did anyone examine the various senses in which bacteria "compute"? Biologists are analogy addicts. They attach weird meanings to words and phrases. Look at Dawkins calling us "survival machines". Many more cases outlined in book related, and that's just in neuroscience!
July 17, 2025 at 2:32 PM
Dubious, and probably also vague! We can only judge justifications of actions and prescriptions with respect to goals. What goals do the editors have in mind? Investigating traits with a known structure of causal relationships between them?
July 16, 2025 at 12:28 PM
Might make sense, but not all changes appear tracked. v2 first image, v3 second image. Look at the titles in bold. Does the color not change for formatted text?
July 16, 2025 at 9:27 AM
I've never seen purple text like this in a paper. How should I interpret it? is it for highlights and summaries?
July 15, 2025 at 11:00 PM
Good news: the paper has a remedy. Bad news: need mostly real data. I don't trust AI companies to get it without causing havoc: theft, fraud, crowdsourcing from desperate people. They'll find some way to make it hurt.
July 1, 2025 at 4:39 PM
Less than 3 hours in. It only took just two somewhat long private messages. Someone really doesn't care for the integrity of large "social media" platforms.
June 24, 2025 at 3:25 PM