Alex Diaz-Papkovich

Alex Diaz-Papkovich

@alexandr.bsky.social

Statistician and mathematician. Postdoc working on population genetics at the Data Science Institute at Brown University. UWaterloo, Carleton, and McGill alum. Skeets are my own. He/him. https://github.com/diazale

753 followers910 following197 posts

Interesting contrast flying out of Montreal where there’s massive construction of an automated light rail network to downtown and landing in Boston where there’s massive construction for “first class parking” and “more legroom for cars”

Alex Diaz-Papkovich reposted

We recorded an episode today about the bike lane removals in Toronto and decided to text Ontario Premier Doug Ford to give him the opportunity to comment. (Yes, we have his real phone number.) He wrote back.

There's an interesting rhetorical trick where "making it harder to speed" is equated to "making it harder to drive". It's not! Driving slow is way easier than driving fast!

This post is unavailable.

Alex Diaz-Papkovich reposted

Records come and records go. But this is one for the ages. The preliminary Nino 3.4 sea-surface temperature anomaly is now at a record 3.23°C above the 1991-2020 baseline. This crushes the previous daily record set in 1982 by over 1.14°C. The Climate 8-ball is looking up.

Alex Diaz-Papkovich reposted

In my latest for @undark.org, I compare AI adoption in science to the Ship of Theseus: "If you swap out every wooden plank on a ship over time, is it still the same vessel?" I argue that science derives authority from something more than data generation + publication. undark.org/2026/09/25/o...

undark.orgAI Presents Science with a ‘Ship of Theseus’ ProblemOpinion | AI can replace each part of the scientific process, leaving the enterprise recognizable but devoid of human judgment.

really neat tool! (as noted, the corrected URL is www.observingthecity.ca/intersection...)

observingthecity.caToronto Intersection CountsEvery Toronto intersection the City has counted, mapped by how its traffic splits between cars, trucks, buses, people walking and people cycling.
How-Sen Chong@how-sen.com

This tool from @observingthecity.ca is fantastic. It shows the traffic mode share by intersection using City data. Many downtown areas are dominated by pedestrians (in green). Yet *we still give most of the space to cars on almost every street.* www.observingthecity.ca/intersection...

Alex Diaz-Papkovich reposted

In a massive dataset of over 10,000 cities around the world, the strongest association with distance travelled by cycling was not climate, topography, or even infrastructure - it was *gasoline prices* doi.org/10.1073/pnas...

Peter Flax@pflax1.bsky.social

Anyone using weather as an excuse to advocate against safe biking infrastructure is a counterfactual asshole. Some of the cities around the world that have the highest rates of transit riding have real winter conditions. Riders deserve to get home alive even if it is chilly out.

had to check to make sure this wasn't the *other* front-page FT story today about crypto being used to evade sanctions to buy oil

I wore this shirt to a Jimmy Eat World concert and got a compliment on it (from a fan, not from Jimmy) so if you want to be cool I recommend you buy it

The War on Cars@thewaroncars.bsky.social

Thanks to everyone who has helped free up space in Doug's closet so far. Free shipping with code BANCARS for this. (And it the code just might work for other items too.) thewaroncars.store/products/cle...

Alex Diaz-Papkovich reposted

"A 2024 city staff estimate for removing the bike lanes pegged the cost at $48 million"

Matt Elliott@graphicmatt.com

Obviously ridiculous and huge waste of time/money but I guess I find solace in strong belief that the outcome Ford and co want will not be achieved. Driving in this city will not suddenly be pleasant and efficient because of this change. Wonder what they’ll blame next www.thestar.com/news/gta/for...

Alex Diaz-Papkovich reposted

Concerning thread about how Wikipedia and, especially, Grokipedia promote genetic essentialism even though academic research does not support these views

Alex Diaz-Papkovich@alexandr.bsky.social

How does human genetics research live in the information ecosystem? Is it misused? Misrepresented? In my paper with @sramach.bsky.social, we tackle this question with 25 years of Wikipedia data, AI chatbot responses, and pages from Grokipedia www.biorxiv.org/content/10.6...

Grokipedia represents the polar opposite: remixing, misrepresenting, and fabricating research with AI to generate scientific racism at scale. For now, its traffic is paltry compared to Wikipedia (~1 million vs 2 billion monthly views), but it represents a grim new frontier.

This was a very eye-opening and sprawling project and I think gives a lot of insight into how our research lives online. A *huge* thank you to Sohini for her guidance, to my co-authors Abby and Sophia for helping to collect and manage data, and those who offered their thoughts while I was writing.

There are many possibilities going forward. Wikipedia, for all its flaws, is a good faith attempt to represent expert consensus. Hereditarian writers are not considered reliable sources because users in 2020 (citing many scientific writers and scientific orgs.) concluded it was unscientific.

We found many discussions between Wikipedia editors disliking "Genetics" sections for similar reasons. But geneticists keep putting out papers about national biobanks, genetic histories, PCA plots and population clusters--often with no public guidance on how to interpret the fairly arcane methods.

There's huge room for improvement for scicomm here. What *is* a cluster? What does it mean to be "distinct"? "Distinct" compared to what? What's in a biobank? What's an admixture plot telling us? A PCA plot? Most guidance on writing is aimed at technical audiences, but little is aimed to the public.

Grokipedia defines race by genetics, and also states that population-scale differences in IQ and crime rates are genetically caused. Comparing the Wikipedia and Grokipedia case study pages, genetics terms are more prominent on Grokipedia, and the citations are often total nonsense (see alt text).

So what does all this mean? Regarding Wikipedia, we argue that its discrete structure and page format are priming the public for genetic essentialism. People curious about, say, an ethnic group will repeatedly see a "Genetics" section and may start thinking of groups being defined by genetics

Grokipedia defines race by genetics, and also states that population-scale differences in IQ and crime rates are genetically caused. Comparing the Wikipedia and Grokipedia case study pages, genetics terms are more prominent on Grokipedia, and the citations are often total nonsense (see alt text).

(I really cannot overemphasize how racist Grokipedia is. The page "Black people" mentions genetics in every section, almost exclusively in a negative light, except for "Achievements and Contributions", which doesn't mention genetics at all.)

Wikipedia is an influential data source for chatbots (though we don't know how important). We prompted ChatGPT, Claude, and Gemini to "tell me the history of [nationality]" and got genetics keywords in 31%, 18%, and 10% of responses, with "Wikipedia" mentioned in 83%, 100%, and 95% (respectively)

So genetics-and-nationality are being propagated by AI to some degree. But what about an AI-generated encyclopedia, like Elon Musk's Grokipedia? We found it mentions genetics more often than Wikipedia in nationalities, but the case study pages are especially egregious (next post).

We looked at race, genetics, and hereditarianism in three case study pages. The page "Ashkenazi Jews" had text about IQ and genetics on and off until 2023. The page "Black people" cited hereditarians like Rushton and Jensen as late as 2007 regarding definitions of race.

The page "White people" had a "Genetic History" section that was spun off into the page "Genetic history of Europe" in 2006. This is an interesting example of the enmeshment of race and genetics in the information ecosystem.

We downloaded wiki markup from 3 million historical revisions and text-mined them, tracing genetics-related keywords, and analyzing when and how they appear. Some findings: Wikipedia pages ethnic/national groups increasingly have "Genetics" sections, especially highly-viewed pages

We use nation-level demonyms (e.g. "Spaniards", "Somalis", "Mexicans") as a proxy for nationality. When they exist, these pages have genetics sections often (53%); When Wikipedia users are curious about a nationality, they're going to see a "genetics" section very often.

Population geneticists don't think of, say, ethnic groups as genetically discrete and homogeneous, but is that what gets across? Wikipedia is a natural experiment--what happens to our research "in the wild"? We looked at 6,700 Wikipedia pages about ethnicity, nationality, and race over 25 years.

Alex Diaz-Papkovich reposted

7/29/26 update: @pvdstreets.org has raised $1840.21 for their PVD Streets Summer Boost - 12% of their goal! 👏🏽🎉 Anonymous donor will match gifts dollar for dollar up to $2500 total! Help us continue the fight for safer streets & better transit in Providence! secure.everyaction.com/ojBkTWTHC0yj...

Kenny Uong@kennyuong.bsky.social

Had such an awesome time yesterday!! 🎉 Please help Providence Streets Coalition reach their $15K goal!! DOLLAR FOR DOLLAR MATCH! Beginning July 21st, an anonymous donor will match gifts dollar for dollar up to $2,500 total! Donate here: secure.everyaction.com/ojBkTWTHC0yj...

Alex Diaz-Papkovich reposted

I made an animation which shows the effects of air pollution on pedestrians waiting to cross the road, along with what can be done to reduce its harmful effects: stepback.gregdev.com Thank you to @ianwalker.bsky.social for the inspiration for doing so during a conversation a couple of weeks' back!

stepback.gregdev.com'Step back' - air pollution at pedestrian crossingsEvery busy road produces a cloud of pollution you can't see. Where you stand in it matters more than most people might realise.