Actually, #ChatGPT is helpful understand how to explain things to people. Long haul, but I got it to admit #Elixir is safer than #Rust if we restrict to the languages (not programmer) and the definition of safety as memory issues, race conditions & undefined behavior. :)
ChatGPT's knowledge of programming languages is often incompetent contradictory gibberish, and scary. As a coder, I will sleep better tonite knowing this AI will probably only get worse as it trains off it's own crowd-generated nonsense. #Elixir and #Java examples.
Is functional programming a lie? (spoiler: no, but mutation is a lie). As usual an #elixirlang promotion in disguise :) - Philosophy, controversy, etc. tomjoro.github.io/2021-12-14…
Today I tried to use a local Python Jupyter Notebook and here are the training times:
0.6s - Elixir Livebook
2.7s - Python 3.9.4
12s - Jupyter Notebook
23s - Google Colab local runtime
1m50s - Google Colab free hosted runtime
I was playing around with Google Colab and Elixir Livebook today. The former required 1m50s to train my model. The latter required 0.6s to train the same model. That was 172x faster! [1/3] #MyElixirStatus#elixir#nx
(after reading the teasers from José and guessing) A few weeks ago I tweeted about how great a match Elixir, with its immutable-functional-concurrency, would be for the GPU. It's coming true - Yay!
In this talk @tomjoro & @m1halo, share their experiences introducing Elixir to OP Financial group. Several production Elixir systems have been launched, including a payment system based on facial recognition. Watch the talk at piped.video/NvyrKcDtpUQ#FinTech#BestOfTheBEAM
In this presentation I pose the question: How does one safely "kill" a thread in #java, #Ruby, #csharp, #rustlang, (or any language with a shared memory model)? How about in #elixirlang or #erlang? Why?
Presenting in front of a "virtual" (but very real) audience is nerve-racking! We need to make on-line presentations more engaging - I need to hear people laugh at the jokes.
While hw people are nitpicking hw performance: #AMD vs Apple/#Arm vs #Intel , vs 7nm, 5nm, etc. -- there are a bunch of sw engineers thinking: welcome to the parallel age, your old sw benchmarks do not apply, and we need to retool software development. #elixirlang, #golang
"To power their chat systems, Discord runs 400-500 Elixir machines. The most impressive feat is that Discord’s chat infrastructure team comprises five engineers. They all learned Elixir on the job."
Geez.
elixir-lang.org/blog/2020/10…