Edu Tuario
edutuario.bsky.social
Edu Tuario
@edutuario.bsky.social
Cronopio cronopio
y dibujitos
Reposted by Edu Tuario
tldr quote:
As demographers who study fertility, family behaviors and childbearing intentions, we can say with certainty that population decline is not imminent, inevitable or necessarily catastrophic.

Love it! Practical, optimistic, hinting at actual problems and solutions (!)
July 29, 2025 at 10:35 AM
Reposted by Edu Tuario
Grok, OpenAI, Anthropic — all the frontier labs are rushing to help the US government

well, except the Chinese labs

www.anthropic.com/news/anthrop...
Anthropic awarded $200M DOD agreement for AI capabilities
The U.S. Department of Defense (DOD), through its Chief Digital and Artificial Intelligence Office (CDAO), has awarded Anthropic a two-year prototype other transaction agreement with a $200 million ce...
www.anthropic.com
July 15, 2025 at 12:56 AM
Reposted by Edu Tuario
Kimi K2, the new best open model in the world. In a few days the non-tech part of US/world is going to have another mini DeepSeek style existential crisis because of this model.

I wrote about my expectations for how this release plays out and what it means for the future.

buff.ly/LhJmcYo
July 14, 2025 at 3:22 PM
Reposted by Edu Tuario
Interesting to see supposed climate change activist Leonardo DiCaprio at Bezos' vulgar wedding.

Worth remembering stuff like that when we are expected to take celebrities and their causes seriously. It's only ever about optics.
June 30, 2025 at 12:47 PM
Reposted by Edu Tuario
This is pretty neat, GAN training in the browser. For toy domains like MNIST and CIFAR-10, but still
GAN playground: Experiment with Generative Adversarial Networks in your browser
reiinakano.com
April 15, 2025 at 11:01 PM
Reposted by Edu Tuario
First a predictive model for identifying future criminals
Then this operational optimization of deportation machinery
I can imagine a microservice on AWS crunching numbers, deciding which neighborhood to send a truck to

A horrible vision of the future world that some ppl have. We need to fight it.
“like (Amazon) Prime, but with human beings.”

michiganadvance.com/2025/04/09/i...
April 10, 2025 at 7:25 AM
Reposted by Edu Tuario
Some time ago, I realized that our existential plight is that we have to take a stand and commit to the values that make life worth living; that one chooses to stand with similar people, and even if this stand is desperate, it’s still the path that freedom has to take, if it wants to be free.
April 7, 2025 at 8:49 PM
Reposted by Edu Tuario
That Crypto is a parasitic entity dependent on a functioning global finance system rather than a revolutionary framework for replacing it is going to occur to a lot of people after it is too late to bail out
April 7, 2025 at 6:31 AM
Reposted by Edu Tuario
3️⃣ Research publications – like the "Unsolved Problems in ML Safety" paper – establish the boundaries of what issues AI safety prioritizes as a field, and are a core avenue through which the field’s ideas have begun to take hold outside the AI safety epistemic community.
October 30, 2023 at 9:47 PM
Reposted by Edu Tuario
Linguistics is everywhere! Who would have thought our field would turn out to be so popular? Law, biology, machine learning, they all want to know about linguistics now.

101 in Language and AI in 101s:
www.youtube.com/watch?v=gccA...
101 of AI and language in 101 seconds | 101 in 101
YouTube video by UC Berkeley
www.youtube.com
March 13, 2025 at 8:18 PM
Reposted by Edu Tuario
Takeaways from the latest State of ML Competitions report mlcontests.com/state-of-mac...:
- Python & PyTorch still dominate
- 80%+ use NVIDIA GPUs, but no multi-node setups 🤔
- LoRA still popular for training efficiency, but full finetuning gains traction.
Surprisingly, CNNs still lead in CV comps
March 5, 2025 at 4:24 PM
Let's see if all of those "free speech warriors" like Jordan Peterson, Weinstein, Joe Rogan et al make a comment on this. USA is really going full fascist.
March 4, 2025 at 4:54 PM
Reposted by Edu Tuario
Here’s the 2025 LLM roadmap 😊
1. Code and train your own LLM to really understand the fundamentals
2. Train models more conveniently using production-ready libraries
3. Learn about the big-picture considerations for real-world LLM/AI apps
February 23, 2025 at 7:41 PM
Reposted by Edu Tuario
After 6+ months in the making and over a year of GPU compute, we're excited to release the "Ultra-Scale Playbook": hf.co/spaces/nanot...

A book to learn all about 5D parallelism, ZeRO, CUDA kernels, how/why overlap compute & coms with theory, motivation, interactive plots and 4000+ experiments!
The Ultra-Scale Playbook - a Hugging Face Space by nanotron
The ultimate guide to training LLM on large GPU Clusters
hf.co
February 19, 2025 at 6:10 PM
Reposted by Edu Tuario
It's 2025, and I’ve finally updated my Python setup guide to use uv + venv instead of conda + pip!
Here's my go-to recommendation for uv + venv in Python projects for faster installs, better dependency management: github.com/rasbt/LLMs-f...
(Any additional suggestions?)
February 15, 2025 at 7:14 PM
Reposted by Edu Tuario
One of the most interesting new research in the LLM space: introducing Large Language Diffusion Models. arxiv.org/pdf/2502.09992
arxiv.org
February 17, 2025 at 9:49 AM
Reposted by Edu Tuario
Review! mini kuš! #127: Swelling – Walker Tate’s Body Horror Short is Both Disconcerting and Bleakly Funny. From @kushkomikss.bsky.social. www.brokenfrontier.com/mini-kus-127/
mini kuš! #127: Swelling - Walker Tate's Body Horror Short is Both Disconcerting and Bleakly Funny – Broken Frontier
A return to the 'mini kuš!' range of minicomics from Latvian publisher kuš! comics as we look at Walker Tate's 'Swelling'.
www.brokenfrontier.com
February 7, 2025 at 4:57 PM
Un liquore che brucia la gola, per scaldare l'anima, abbiamo un sangue molto caldo che spaventa i fantasmi xenofobi
January 24, 2025 at 11:27 AM
Reposted by Edu Tuario
I’ve put a limited run of this A2 print on my shop, 50% of all sales will go to @medicalaidpal.bsky.social . There won’t be a reprint of it so when they’re gone that’s all there is. Thanks all.

wefail.art
December 7, 2024 at 9:28 AM