joaogui1

AlphaCode is solving much harder problems than these first few days of AoC

What's TFA?

What difficulties did you face using Coq?

Not just DDG, there has been an increase in search engine startups, some with pretty well-known people taking the lead, that probably got them worried

TL;DR: Homa implements a novel and effective form of bipartite matching that allows it to achieve high network utilization while also respecting message priorities.

Julia is more for Scientific Computing than Data Science, think weather prediction and fluid simulations vs data visualization and gradient boosted trees. In that niche Julia is doing pretty well in fact. Also it's an interesting language in general, with a nice macro system and a powerful multiple dispatch one, it's definitely worth learning IMO.

It has friction because you're not working in a vacuum, pun intended. The more people you're working with, the more friction/overhead there is. The trick is trying to automate as much of it as possible, and to batch what's impossible to automate

Julia has this

That guy is a famous Kaggler that works for Nvidia, not a minor Twitter user

This would be amazing, please try releasing some of these blog posts :D

Wasn't there a reduction in suicides? Not that that's the only factor when it comes to mental health, but it's probably something to take into account when you talk about the mental health toll of the past 2 years

Except for LeCun that defends Facebook even more than their PR team (denying Facebook's use in the Rohingya genocide even after FB apologized)

It's not a vanity division, but I would definitely put Facebook in fourth, below Google Research, DeepMind and Microsoft, so definely not safely on second place

Christopher Alexander was the author of A Pattern Language, The Timeless Way of Building, and many other architecture books. His work has influenced Computer Science through the idea of design patterns, games by being the main inspiration for SimCity and also influencing Spore, among many other contributions. Here's a tweet from an account dedicated to his book a pattern language about his passing: https://twitter.com/apatterntolearn/status/15048528809641328... Definetely one of the giants of the last century, may his memory be a blessing.

There is much more rationalising about death being good in Christianity (with the notion of a heaven for the believers) than in Judaism (no set beliefs on the after life) or even many pagan religions of that time (Greek life after death is often depressing for non-heroes)

Idris, some Haskell extension, and Agda

I think that means Mastodon also doesn't count as a social media (not disagreeing)

There is an initial version of Roc, though the repo is private and you need to ask Richard Feldman for access

Aspergers/ASD is a neurological condition, not just psychological

Oooh, didn't know that, cool!

Which one?

Wait is this Elixir?

ADA is the company doing chatbots/conversational AI, cohere is providing them with language models to do it, that's not the same thing as cohere being a company in the chatbot field

I'll just say that the word democratize does not appear in Cohere's website

Any idea what are the specs?

Did you use orgmode and mu4e? If so what are you using in their place now?

They breath oxygen, they "feed" on CO2, once fully grown they need less food and that changes the balance

While jax has those operations you don't always need them, it depends on what transformations you want to do (JIT or grad) and they have been working on making normal control structures compatible with all transformations

In your first sentence you're mistaking JAX and XLA

XLA: Accelerated Linear Algebra, I guess it's kind of a backend/compiler that optimizes Linear Algebra/Deep Learning calculations with some very interesting techniques, among them fusing kernels

JAX: In some sense syntax sugar over XLA, but a better way of describing it is Composable transformations + Numpy + some Scipy. The composable transformations allow you to take derivatives (be them single, multi or vector valued functions and also higher order derivatives), JIT a function (which is them compiled to XLA), 2 forms of parallelism (vmap and pmap) and others, while being compatible with one another and with both TPUs, GPUs and CPUs

There are a bunch of frameworks built on top of Pytorch too (fastAI, lighting, torchbearer, ignite...), I don't see why this should be a problem (or at least a problem to JAX but not to Pytorch)