AZ native, son of immigrants, ex-Navy PhD Social Psychology – dissertation on envy and antisemitism Data science, media reliability, bias in academic research
Arizona
Joined July 2014
- Tweets51.3K
- Following3.1K
- Followers5.4K
- Likes95.1K
Political Diversity Will Improve Psychological Science
cambridge.org/core/journals/…
Jyrki's latest bake.
Fascinating project on acoustic similarity.
All of Jyrki's software projects are named after Swiss German baked goods. Brotli, the web compression codec, is a small bread roll, Zimtohrli is a cinnamon ear, etc.
Q: "How similar would two wav files sound to a human?"
A: Zimtohrli : An efficient psychoacoustic audio similarity metric
Zimtohrli has a 128-channel equivalent rectangular bandwidth-spaced gammatone, a modified dynamic time warping and a neurogram similarity index, all optimized against human ratings. Best OSS audio metric. Best for full-band music, even better than commercial metrics.
pubs.aip.org/asa/jel/article…
They're not even extreme statements. Banning abortion? Trans as mental illness? This is a very confusing study and result.
I just watched the stupid "Who's on first" skit for the first time because of the Holmes thing.
It's not funny, or about as funny as Seinfeld's stand-up intros before the Seinfeld episodes.
I've been confused by filmmaking ever since I learned about editors and the editing process. If the director doesn't control the edit, how can the film be true to their vision?
Or at least, how can films be consistently true to the director's vision? How is a director a filmmaker, in full, if someone else does the editing?
I wish they provided context to explain what they're talking about. I don't understand what it means to use a diversity survey to cover for a planned investigation, or the implications.
Do you anticipate that lots of Washington doctors will be willing to lie about a child's sex? I see that as a potential weakness here, esp compared to the easy cheek swab the pros use.
The "Science Only" section in my piece at @wesyang's Year Zero covers a related issue well – the lack of evidence*. The four steps of asking whether a given artifact of "evidence" is actually a study, that it covers minors getting the monstrous ministrations, etc. whittles down to zero if we require the studies be valid (randomized, long-term follow-up, no suppressed outcome variables like Olson-Kennedy did, etc).
Definitely zero if we want to see the data.
* Evidence is a sideshow anyway. We cannot sterilize or mutilate children. It's ethically impossible, especially on cult leftist grounds of "gender identity".
open.substack.com/pub/wesley…
I spawned one to find me a desk chair, given certain criteria. It worked well.
I spawned another one to find images suitable as featured images for my old blog posts (which have no images at all), based on the topics of the posts. This didn't work as well, though it's not clear what kind of image fits with posts on fraudulent social psychology research, invalid and fraudulent climate consensus research*, issues of statistical validity, etc.
* The viral 97% study counted irrelevant papers (they shouldn't be counting any papers to measure consensus) like a welding paper by Volvo engineers, so discussing with Claude got me images of the actual Volvo factory using the new steel, so I might use that.
I wish they had built that clean sheet OS they were building with that modified "System C#" language.
We need clean, much better OSes. It's ridiculous that in 2026 we don't have instant computing on desktop and mobile, where everything takes a fraction of a second (except transcoding video, etc.) Apps should open in 150 ms or less, fully ready for input. And a profoundly higher level of security should be assured in any modern OS – none of the memory bugs should be possible.
It's crazy how slow computers are, that it takes several seconds to open an app, that almost all websites are very slow, taking seconds to load because of the norm of incompetent web developers who force users to download a bunch of different files to read a 15 kb article.
Microsoft could've been leaders. Instead we're stuck in this weird duopoly with slow computers.
Slapping his phone out of his hand ruined it. It makes you the only one who used force. If there's no recording of the threats, then all the police have is you committing assault.
Joe Duarte retweeted
‘Weak leadership combined with ideological capture, perceived prestige for the organisation, and fear had all led to children being harmed.’
Gender ideology has spread like wildfire through the NHS. The casualties: kids and whistleblowers.
Free to read.
thetimes.com/uk/healthcare/a…
It would be interesting to do an interview on this, on why you think this would change a conservative's pro-enforcement policy stance. You might be assuming that if anyone one cares about suffers, that means the policy is bad, but that wouldn't be a typical conservative view.
The fact that this was some kind of voluntary or uncontested removal makes it a weaker example (and he had no application in play). But maybe there are better, more poignant examples.
Oy vey. It really is bleak.
At Yahoo, Mayer overspent on a company Christmas party where she held court like a queen and Yahoo employees took turns getting their picture with her highness.
Joe Duarte retweeted
🚨Today, Texas Children’s Hospital has officially launched their Detransition Clinic 🚨
Under the settlement, TCH must provide multidisciplinary detransition-related care FREE OF CHARGE for five years, including endocrinology, surgery, primary care, fertility counseling, mental health services, social work, and speech pathology.
I extend my heartfelt gratitude once again to Texas Attorney General Ken Paxton and my attorneys at the Burke Law Group for their extraordinary dedication, sacrifice, and tireless efforts throughout this fight. Their commitment helped turn a painful and difficult chapter into an opportunity for meaningful change. @KenPaxtonTX @BurkeLawTweets
I think of all the detransitioners who have suffered in silence, been dismissed, and fought desperately to have their voices heard. This win is for them.
I hope TCH’s admirable effort to fund and launch Americas first detransition clinic becomes a turning point for the entire country. That this clinic is not the exception, but the first of many.
Today’s a good day. ❤️
Link⬇️
I agree that surveys are invalid. If Ofcom was reporting surveys though, they should've reported it.
Much like Wikipedia with their arbitrary "Reliable Sources" list, Ofcom is trying to just *assert* that leftist media is more reliable than non-leftist media. Then use that as a basis for social media censorship.
They have no evidence for such assertions. There's no systematic study of media reliability (and Ofcom could not be trusted to conduct a valid study).
They had been reporting surveys of viewers, which are not valid measures of trustworthiness, but when the survey data showed that GB News is more trusted by its viewers than the BBC was by its viewers, Ofcom stopped reporting it...
GB News has accused Ofcom of suppressing data showing that its viewers rate the channel as more trustworthy than the BBC, and has threatened to escalate its complaint.
The row comes as the Government presses ahead with plans to make social media platforms promote “trustworthy” news sources such as the BBC, ITV and Channel 4. The proposals could see more critical outlets, including GB News and The Telegraph, demoted.
Ministers have yet to explain clearly how broadcasters and newspapers would qualify as “trustworthy”.
An Ofcom survey found that GB News viewers rated the channel more highly for trustworthiness than viewers of the BBC, Sky News, ITV and Channel 4 rated those broadcasters. GB News viewers also rated it above all other British broadcasters for the range of viewpoints on offer, and joint first with Sky for impartiality.
Ofcom published trustworthiness figures in previous years but omitted them from its 2026 report. GB News has questioned whether the regulator deliberately withheld evidence about the channel.
Ofcom later blamed an administrative error. It also said it had left out the figures because a change in methodology meant comparisons with previous years could be misleading.
@GBNEWS chief executive Angelos Frangopoulos has raised concerns that Ofcom treated other broadcasters differently. He has asked the regulator to hand over correspondence with Culture Secretary Lisa Nandy and said he would contact her if it failed to do so.
In a letter to Ofcom, Mr Frangopoulos wrote: “It is difficult to avoid drawing a conclusion that Ofcom is failing to apply a fair, consistent and transparent approach to its own research findings across broadcasters.”
Read more below 👇
