— Как вам удалось так похудеть? В чём ваш секрет?
— Перестал жрать говно в Твиттере...

Видишь в итальянской пасте укроп с петрушкой, и невольно думаешь "откуда тут африканские названия?!"

I recently read a prescient short fiction story "See no evil" by John R. Pierce, where he anticipated in great detail the blueprint for most modern AI startups, already 55 years ago!

Unfortunately the only reference to it I could find is this one: https://www.isfdb.org/cgi-bin/title.cgi?68334

http://aria.org.uk/wp-content/uploads/2023/10/ARIA-Mathematics-and-modelling-are-the-keys-we-need-to-safely-unlock-transformative-AI-v01.pdf
"Mathematics and modelling are the keys we need to safely unlock transformative AI"

David “davidad” Dalrymple outlines the case for mathematical "world models" and formal verification as a fruitful or even necessary approach to AI alignment and safety.

The 3-page PDF features very nice layout and a long list of relevant publications (including deep technical ones) for further study.

But the best part: ARIA (aka Advanced Research and Invention Agency) is going to provide funding for research in this area, and starts to collect (preliminary) research projects proposals.

If you have a question and think to yourself "I'll ask a LLM about it!" better ask two of them and compare the answers providing your feedback to the community:

https://chat.lmsys.org/

😊

— Всё, чтобы вы отъебались улыбались!

Моему коту в этой жизни нужно только одно — укрытие от жизни.

https://www.stochasticlifestyle.com/chatgpt-performs-better-on-julia-than-python-and-r-for-large-language-model-llm-code-generation-why/

Evidently ChatGPT 3.5 "understands" #Julia significantly better than other languages including Python and JavaScript, not even mentioning Go, C and C++: https://arxiv.org/abs/2308.04477

Chris Rackauckas gives some points as to why it's not that surprising from the perspective of teaching novices. And also speculates that large volume of so-so tutorials and examples for very popular languages might hurt LLMs' learning.

He also advertises his diffeqpy library connecting optimized solvers implemented in Julia (including code generation for GPUs) to #Python #machinelearning libraries, and points to some curious papers with impressive benchmarks. 😊

Хотел сократить "психологический уют" до "псих. уют", но получилось как-то неблагозвучно... 😏

https://www.icms.org.uk/funding-opportunities/mathematics-humanity

Turns out there's an International Centre for Mathematical Sciences in the Edinburgh (which is not very surprising at all by itself), and there's a Mathematics for Humanity program inside this centre.

The program is devoted to three major tracks:
— Integrating the global research community (including educational projects)
— Mathematical challenges for humanity (epidemiology, climate modeling, energy systems and even AI alignment)
— Global history of mathematics

The best part: they accept proposals for projects until 1 Jun 2024 and will fund the selected ones.

https://tshort.github.io/WebAssemblyCompiler.jl/stable/

Julia too can now run in a Web browser (or another Wasm runtime) some subset of programs! 😁

Major highlights:
— uses Binaryen to compile Julia's IR to WebAssembly
— uses Wasm-GC (Garbage Collector proposal)
— can interoperate with JavaScript and access DOM
— can use Plotly.js 😃

Current limitations include:
— no multi-dimesional arrays
— no pointers (arguably a good thing 😏)
— no union types
— no exceptions
— no BLAS nor other C functions

Yeah, that's only a Proof-of-Concept, though quite capable already (can run Julia-native ODE solvers from DifferentialEquations.jl). 💪

https://blog.arxiv.org/2023/12/21/accessibility-update-arxiv-now-offers-papers-in-html-format/

Our favorite blogging platform — arXiv — now provides an automatic conversion to HTML to improve accessibility for the readers!

https://marketplace.visualstudio.com/items?itemName=paolomasci.vscode-pvs

WOW! There's a VS Code plugin for the PVS theorem prover. From NASA of course.

https://dmkpress.com/catalog/computer/programming/978-5-93700-234-1/

В ДМК-пресс выходит перевод знаменитой The Little Schemer с подзаголовком "чудесное функциональное программирование". Как перевести заголовок в этот раз не придумали, но судя по прошлым опытам — оно только к лучшему. 😏

А я сяду в кобылолет...

https://johncarlosbaez.wordpress.com/2023/11/25/what-can-mathematicians-do-about-climate-change/

John Baez compiled a short and somewhat superficial list of abstract mathematical fields and broad topics that nevertheless useful for addressing Climate Change, especially in the area of modeling human response and human actions in general.

Also interesting links in the comments...

Мы смотрим в будущее с непоколебимым астигматизмом!

https://undonecs.sciencesconf.org/resource/page/id/11
1st conference on Undone Science in Computer Science
"A conference to reflect on epistemological and ethical dimensions of computer science"
February 5-7 2024

Accepted talks cover a broad range of topics: data science and AI, healthcare, education, formal specifications, proofs and tests, and even the nature of Theoretical Computer Science itself.

They promise free online translation, but I'm still to find the URL for it...

https://www.youtube.com/watch?v=T0g66FbVaow

Researchers from Stanford (yeah, what's about Google there?) built a mobile "bimanual" (two "arms"/manipulators) robotic platform from readily-available relatively inexpensive (~$32k total, about an order of magnitude cheaper than commercial offerings) consumer-grade hardware. And they open-sourced all the code, schematics, parts and build instructions.

I'm personally not that excited about "a personal house robot that will do my dishes" as of yet, for one, you have to directly manipulate the platform — significant number of times! — for it to "pick up" a skill and reproduce the same movements. You can't simply tell it "cook me a dinner", you have to actually cook it yourself several times using robot's manipulators. At least you can't for now. 😊

Anyway, very impressive achievement, a real showcase that research and industry haven't stalled, and we're getting closer to useful consumer robots.

https://www.theguardian.com/environment/2023/dec/22/zombie-deer-disease-yellowstone-scientists-fears-fatal-chronic-wasting-disease-cwd-jump-species-barrier-humans-aoe

"Zombie deer"! Though from Yellowstone and not South Korea. Yet. 😏

Actually it's a case of ongoing growing epidemic of Chronic Wasting Disease which is caused by prions — misfolded proteins. Which is fascinating — it's not an organism in any sense, not a parasite, bacteria or fungi, not even a "half-organism" like a virus that at least carries an RNA and reproduces through host's "DNA-copying" machinery. It's just an organic molecule. Even kinda broken one. But that's exactly the catch: as long as it's an "innocent protein" an organism is pretty "welcoming" to it, and as long as it's broken it can cause a "chain-reaction" of broken proteins.

The bottom line — it causes a 100% fatal encephalopathy in deer, cattle and humans. It seems the first human Creutzfeldt–Jakob disease pandemic is going to be the last one... 😏

https://reasonablypolymorphic.com/blog/coblub/index.html

Sandy Maguire on Agda and co-(Blub paradox). 😊