Heads up that Manning is having a half-off books sale. They have a lot of good data science titles. I was interviewed for this one 😊 www.manning.com/books/build-...

If you ever see me post non-data-related stuff here, it's because I forgot that I switched accounts 😅 If you're interested in following my personal account, it's @paix120.bsky.social

🤔 @becomingdatasci.bsky.social new way to learn SQL just dropped

Looking forward to reading this

Here's an Epidemiology starter pack! go.bsky.app/K6DXCGi

Apparently there are a bunch of new people coming over from Twitter... Drop your data science starter packs in the replies for people to follow!

Some of my favs: bsky.app/profile/did:...

Two "data people" ones: 1: bsky.app/starter-pack... 2: bsky.app/starter-pack... and this is where people can search starter packs (my go-to resource for those new to Bluesky!): blueskydirectory.com/starter-packs

Apparently there are a bunch of new people coming over from Twitter... Drop your data science starter packs in the replies for people to follow!

I would love to see the AI skill training that Accenture used. Anyone interested in leaking it to me? www.cnbc.com/2025/09/26/a...

No.

Who are the best voices in the "future, given AI" analysis space? Like,writing about a potential point at which there isn't enough publicly-available human-generated raw content to train them on. Or where there aren't enough experienced staff for a role because AI impacted the workforce pipeline

Lazy post since I haven't yet searched for an answer: How would you determine if you're good at prompt engineering? Are there tests? Are there categories, like you can hone skills for a specific type of Gen AI use case, or specific models?

anybody got a good <1B?

Lazy post since I haven't yet searched for an answer: How would you determine if you're good at prompt engineering? Are there tests? Are there categories, like you can hone skills for a specific type of Gen AI use case, or specific models?

I did the first of two of these yesterday, and more than 50 early career scientists from at least 5 continents joined us! The second is tomorrow afternoon, and there's still time to sign up

Warning to those who work with people's private data that this is the stuff of nightmares Gift link

In a first, Google has released data on how much energy an AI prompt uses www.technologyreview.com/2025/08/21/1...

many benchmarks used to measure AI capabilities are, I think, contrived and lenient. here's a good real-life study, on whether AI can do your (US) tax returns; a domain with plentiful training data and documentation. the result: the best model only got 33% of returns correct arxiv.org/pdf/2507.16126

Same with Data Science, and understanding the aspects of the work that goes beyond training models.

I agree

I agree

Bookshop apparently has free shipping right now, a good time to pick up my book and others! 😊 SQL for Data Scientists bookshop.org/p/books/sql-...

The Senate parliamentarian is asking the Senate Commerce Committee to rework its 10-year moratorium on enforcing state AI laws.

Dystopian.

A peek into how our federal government is using AI, which the House just voted against regulating for 10 years.

The tax bill that passed the US House overnight has a clause that prevents state regulation of AI for 10 years 5calls.org/issue/ai-reg...

SCOOP from me: the mass exodus of employees from NOAA is decimating a tiny office that literally keeps our measurements of latitude and longitude accurate — and, former employees tell me, endangers the science of how the US understands global measurements:

I've been watching a bunch of these veo 3 generated video clips and they're scary-realistic.

As others have pointed out, this could stop states from regulating technology like deepfakes and self-driving cars. Also things that don't even exist, yet. www.404media.co/republicans-...

My take is "HELL NO" to LLMs' "potential to help make decisions in high-stakes situations" (they're talking REAL high stakes life or death war or peace scenarios, folks) And the way this is covered, especially the "hawks vs doves" analysis is deeply concerning. What are we doing??

Oh.

Today!

Tomorrow is Independent Bookstore Day! Ignore the Amazon sale (they allow knock-offs that are worse quality and don't benefit the author to sell alongside official books, anyway) and try Bookshop instead. I link to it on my book website: sqlfordatascientists.com/buy-the-book/

Good job team America www.nbcnews.com/news/us-news...

So Google's latest consent violation is them generating "AI" audio of your voice in YouTube as an opt-out "service." Here's how to turn it off (original images reposted with alt text):

Tutorials day at #PyDataVirginia2025

I enjoyed meeting many of you today at #pydatavirginia2025!

If you're looking for me at PyData Virginia today, here's how I look 😊

I'll be speaking at PyData Virginia next Friday on a panel after lunch about "real world" data scientist skills, and how to stand out on the job and at interviews. Who's joining us in Charlottesville? Also, check out the awesome tutorials for Saturday. pydata.org/virginia2025...

😡

I'll be speaking at PyData Virginia next Friday on a panel after lunch about "real world" data scientist skills, and how to stand out on the job and at interviews. Who's joining us in Charlottesville? Also, check out the awesome tutorials for Saturday. pydata.org/virginia2025...

The article also mentions that DOGE is illegally not preserving their own communications. I wonder if the people who built Grok anticipated this use case.

As a person who works with data and cares about people, I would personally never work with or for Palantir. I assume they already know that about me, though.

Reminds me of our #craftydataviz competition from back in the Twitter days 😁

@wesmckinney.com Glad to have found you on here so I could share this 😆 www.instagram.com/reel/DIFlR3z...

I'm a data person. This is very bad.

Does anyone have info about the UCI ML repository being down? I know someone who was looking to use it and noticed today it doesn't load. archive.ics.uci.edu/datasets