Chris Elrod

[Chris Elrod @BayesWatch]

@Viral_B_Shah @github @github nothing communicated from GitHub, and googling the message brings up only two hits from foreign websites, so I am 100% in the dark.

[Chad Scherrer @ChadScherrer]

UNICODE NEEDS BETTER SUBSCRIPT/SUPERSCRIPT COVERAGE

https://t.co/TuE31994Fp

Easy ways to help make this happen:
• Retweet this
• Star the repo
• Show it to others

[Chris Elrod @BayesWatch]

Preach: https://t.co/4RfrGaCvIq

[Ethan Mollick @emollick]

Tired: startups using machine learning to detect cancer in scans with 90% accuracy.

Wired: startups using flocks of pigeons to detect cancer in scans, with 99% accuracy. https://t.co/W7rWW31EPD

[Chris Elrod @BayesWatch]

@Viral_B_Shah @github @github nothing communicated from GitHub, and googling the message brings up only two hits from foreign websites, so I am 100% in the dark.

[Chris Elrod @BayesWatch]

@github
disabled all GitHub Actions without explanation for:
https://t.co/gpXi2h8UJn
https://t.co/EHbqBr6Et2
https://t.co/Zc1t3LoDNU
This includes CI and Julia TagBot) meaning I have no automated testing nor can I easily create releases so users have access to bug fixes.

[Chris Elrod @BayesWatch]

@github still no word about disabling GitHub actions for multiple repositories.

[Viral B. Shah @Viral_B_Shah]

#Julialang 1.7-dev now allows selecting your BLAS at runtime on all OSes. All you need to do is `using MKL`. All made possible due to @staticfloat 's amazing work on libblastrampoline: https://t.co/6LdXe2TFKx

[Viral B. Shah @Viral_B_Shah]

@BayesWatch I would have thought that if this was the case, it would have been communicated by @github. Curious to see if it is like the internal spam classification system issue like in https://t.co/Zt6c9tvpMk or over-usage of CI that was not communicated, or perhaps something else.

[Chris Elrod @BayesWatch]

I've filed a ticket, but no response yet.

[Chris Elrod @BayesWatch]

I assume this is because LoopVectorization was using a lot of CI (would have been less if comits auto-canceled old runs), and the other two packages happened to have similar names.

[Viral B. Shah @Viral_B_Shah]

Even as #julialang 1.7 makes it easy to replace OpenBLAS with MKL, the community keeps pushing on native-Julia BLAS. Love this snippet from the Octavian.jl README: https://t.co/oogIeQbdiX

[Chris Elrod @BayesWatch]

@markkitti @tomaklutfu @tischitischer @AvX Thanks for chiming in on Zulip! While

> there's a gigantic amount of different CPU architectures out there and the idea of being able to be the best at everything on all of them seems kinda crazy
- @mason_protter

That's the goal, and folks sharing their results help!

[El Mahdi El Mhamdi @L_badikho]

Except a few who spoke up, the AI community has failed in stopping the irresponsible premature commercialisation of GPT3.
It's an overdue tweet, I was busy with faculty job negotiations and a side work on online harassment…
But this matters a lot as it set a dangerous precedent.

[Chris Elrod @BayesWatch]

@sir_deenicus @ivn_finaev @ChadScherrer @l_II_llI Trump said about Nixon:
"He left. I don’t leave. A big difference."
I don't know the context of the 2014 tweet, so I'm liable to be wrong, but I don't think he's making a comment about being gracious in defeat. He thinks anyone who accepts it at all is a loser.

[Chris Elrod @BayesWatch]

https://t.co/glmuzf4WDX

[Chris Elrod @BayesWatch]

@Viral_B_Shah @MoseGiordano @nassarhuda Thank you so much! I'm excited for all that we can accomplish together and to see Julia lead the cutting edge of performance across scientific applications.

[Chris Elrod @BayesWatch]

I am honored by this award, and excited to be a part of this amazing #JuliaLang community. Congratulations to @MoseGiordano and @nassarhuda!

[Viral B. Shah @Viral_B_Shah]

The #JuliaLang community prizes were just announced. All amazing contributors: @MoseGiordano, @nassarhuda and Chris Elrod. Phenomenal contributions from all of them.

[Viral B. Shah @Viral_B_Shah]

The #JuliaLang community prizes were just announced. All amazing contributors: @MoseGiordano, @nassarhuda and Chris Elrod. Phenomenal contributions from all of them.

[Chris Elrod @BayesWatch]

which is why I implemented a random number generator that takes advantage of modern hardware to be fast: https://t.co/HovkJ5VSaq

[Chris Elrod @BayesWatch]

"more than half of the computing time of today’s super- computers is used just for the generation of random numbers." -Gerstner & Griebel