Mastodawn

Björn Apr 10, 2023

Hat tip to this post on the birdsite: https://twitter.com/MattBoisgontier/status/1645114817278091273?t=gYXufPXfuFMPD8BBrX0o9g&s=19

Matthieu Boisgontier on Twitter

“93% of authors indicating their data is available upon request do not respond or refuse to share their data. https://t.co/eVQCGhRlgN”

Twitter

Björn Apr 10, 2023

This paper very much matches my personal experience. I'm personally still at 0% for obtaining data that "are available upon (reasonable) request". Of course, then there's the all data are directly displayed on the tables/figures, when they are clearly not...
https://doi.org/10.1016/j.jclinepi.2022.05.019

Björn Apr 7, 2023

Jenny Chase Apr 7, 2023

Reading "Beyond Measure: the Hidden History of Measurement" and it's really interesting how old ploughing units like the German Tagwerk (& maybe originally the acre?) included information on how hard the land was to plow, so didn't actually translate straight to hectares.

Modern me: this is super annoying.

Me as German peasant farmer: no this makes perfect sense for what I want to use it for.

Björn Apr 6, 2023

Frank Weber Apr 6, 2023

projpred 2.5.0 has been released on CRAN. Most noteworthy user-facing changes are an enhanced verbose mode (extensible by a new global option, thanks to Juho Timonen), a revised "Troubleshooting" section in the main vignette, and minor enhancements in the `print.vselsummary()` output. For details, see the changelog: https://mc-stan.org/projpred/news/index.html#projpred-250

Changelog

Björn Apr 2, 2023

April 1 once more drove home the difficulty of telling apart parody and over-the-top enthusiasm about AI. Watch out, unless your April fools claims that ChatGPT has achieved consciousness, that with large N you don't need randomisation or that you should bomb GPU clusters to stop AI, you run the risk that someone might take you seriously.

Björn Mar 29, 2023

Mark Riedl Mar 29, 2023

It’s happening. People are using GPT to review academic papers https://www.linkedin.com/feed/update/urn:li:share:7046083155149103105

Robin Bauwens on LinkedIn: A reviewer rejected my paper, and instead suggested me to familiarize… | 101 comments

A reviewer rejected my paper, and instead suggested me to familiarize myself with the following readings. I could not find them anywhere. After a control in… | 101 comments on LinkedIn

Björn Mar 29, 2023

Janelle Shane Mar 29, 2023

Remember seeing something about GPT-4 doing well on standardized tests? It turns out it may have memorized the answers.
https://aisnakeoil.substack.com/p/gpt-4-and-professional-benchmarks
#gpt4 #AIHype #ThisIsWhyWeDontTestOnTheTrainingData

GPT-4 and professional benchmarks: the wrong answer to the wrong question

OpenAI may have tested on the training data. Besides, human benchmarks are meaningless for bots.

AI Snake Oil

Björn Mar 14, 2023

Mark Riedl Mar 14, 2023

"Open" AI

Björn Mar 12, 2023

Autumm Mar 12, 2023

Wait wait wait wait

Let me get this straight...

LinkedIn wants to generate crappy AI content and then invite me to fix it, for free, under some guise of flattery calling out my "expertise"

Really?

Björn Mar 11, 2023

Jenny Chase Mar 11, 2023

I wish I could just avoid writing about the US solar market in this book. It's complicated and I can't believe it works the way it appears to work. But sadly it's almost half the size of Europe's and a fifth the size of China's, I suppose I have to mention it.