Ramsés V. Morales

Computer Scientist, Musician, Free Speech Advocate. I conjure abstractions out of nothing.

[Jensen Huang @JensenHuang]

For my first post, I’m sharing a letter @NVIDIA signed on why open models matter.

AI will transform every industry, power every company, and be built by every country.

Open models strengthen safety and cybersecurity, accelerate innovation and diffusion, and enable sovereignty.

The world needs both frontier closed models and frontier open models.

https://t.co/AUKzoQ5Ikb

[Ramsés V. Morales @rmorale]

🚀🚀🚀

[friendliai @friendliai]

Welcome to FriendliAI, Brian Cho! 🎉

Joining as Director of Engineering with 6 years at @Meta and 4 at @AptosLabs under his belt — right as our engineering team scales to match the pace of the agent era.🚀

[Charly Wargnier @DataChaz]

The fact that Agatha Christie had access to ChatGPT is undeniable proof that time travel is real

[Greg Lukianoff @glukianoff]

I’ve always found people who bristle at “American exceptionalism” kind of… weird. Not because I lack self-awareness — I’ve spent my career cataloging every way this country fails to live up to its own rules. But that’s exactly why I love it so damn much. We built a system designed to be shamed by its own founding documents, and it still delivered one of the most spectacular, world-altering runs in human history. A genuine force for human flourishing.

I also found the argument against American exceptionalism to be historically illiterate. Here’s a sample of what we were first at:

• The first large-scale democratic republic in human history — not a city-state, not a monarchy with a parliament bolted on, but a bold continental experiment in self-rule, popular sovereignty, and ordered liberty.

• A written Constitution (1789) with separation of powers and checks & balances — still the oldest national constitution in force anywhere.

• The Bill of Rights (1791): the first time a nation wrote “the government cannot touch these” into supreme law and actually meant it. A dare the world copied — from later rights charters to the Universal Declaration of Human Rights.

• Public land-grant universities and mass higher education (Morrill Act), opening college to ordinary people no aristocracy would have let near the gates. (but don’t get me started about what happened after we started. Massively federally funding it.)

• Kitty Hawk, 1903 — first controlled powered flight.

• The Moon, 1969 — still the only ones who’ve been there.

• The world’s largest economy since ~1890, powering unprecedented prosperity through grit and genius.

• The assembly line, skyscraper, transistor, personal computer, ARPANET — the backbone of the modern world.

• Telephone, phonograph, GPS — connecting and powering daily life.

• Surgical anesthesia, polio vaccine — saving and transforming millions of lives.

• Jazz, blues, rock ‘n’ roll — brand new American art forms that conquered the globe.

• Hollywood’s dreams, blue jeans, bourbon, and a culture so open a kid like me could devour sushi, burritos, stuffed cabbage, and tabouli in the same week and rightfully think of it all as American.

That’s the part that fills me with genuine love and pride: not just the power or the wins, but the appetite for freedom, creativity, and reinvention. The audacity to say “We the People” and keep trying to live up to it.

What do you love most about this truly exceptional country? 🇺🇸

[alphaXiv @askalphaxiv]

As believers of open research, we are disappointed to see Anthropic silently degrading Fable 5 for AI development

"Any topic related to building pretraining pipelines, distributed training infrastructure, or ML accelerator design... may have limited effectiveness through Claude via methods such as prompt modification, steering vectors, or parameter-efficient fine-tuning."

Not only do they get to decide what you use LLMs for in research, but this also enables them to silently intervene in your research without you knowing.

This sets a dangerous precedent. If a model refuses openly, users can understand the boundary. If a model falls back to another model, users can still evaluate the difference. But if a model silently modifies or weakens its own answers while still pretending to help, researchers lose the ability to know whether a failed result came from their own idea, their implementation, or an invisible intervention by the model provider.

That is not safety. Safety policies should be transparent, auditable, and user-visible.

On top of that, the people most harmed by this are not the largest labs with massive teams and proprietary infrastructure. It is the independent researchers, academic groups, startups, and open-source builders who rely on public tools to compete, innovate, and pioneer AI for everyone else.

[antirez @antirez]

Welcome to DS4, a specialized inference engine for DeepSeek v4 Flash. https://t.co/UrUJz5I2R1

This project would have been impossible without the existence of llama.cpp and GGML and the work of @ggerganov and all the other contributors. Thanks!

[Nick Levine @status_effects]

New work with @AlecRad and @DavidDuvenaud:

Have you ever dreamed of talking to someone from the past? Introducing talkie, a 13B model trained only on pre-1931 text.

Vintage models should help us to understand how LMs generalize (e.g., can we teach talkie to code?). Thread:

[Ramsés V. Morales @rmorale]

🚀❤️🚀

[Reid Wiseman @astro_reid]

There are no words.

[NASA Earth @NASAEarth]

That's us! 🌍

The Artemis II crew captured beautiful, high-resolution images of our home planet during their journey to the Moon. As @Astro_Christina put it: "You guys look great."

[Ramsés V. Morales @rmorale]

🚀❤️🚀

[delian @zebulgar]

The coolest orbital animation I've seen of Artemis 2

Just really shows you how far away they're flying today and also how precise they need to be to go to the moon

[Ramsés V. Morales @rmorale]

Maestro @jasonelibecker is my @neuralink wish.

[Jason Becker @jasonelibecker]

Jason Becker - "Crip Tone Night" (Official Video)
From Strawberry Jams!
https://t.co/M8ZQ0kOjEk

[Stephen R. C. Hicks @SRCHicks]

Where you go to college matters less. What you get your degree in matters more.

[Ramsés V. Morales @rmorale]

@Noahpinion G!

Don't forget that some of those places have adopted many immigrants from D.

[Ramsés V. Morales @rmorale]

@Mayer Antes de intentar eso, hay que resolver un problema más importante, que previene inversión extrangera: "Rule of Law."

Un inversionista necesita saber "las reglas del juego", y necesita saber que esas reglas van a ser respetadas por todos.

[Ramsés V. Morales @rmorale]

@kmele It is sad how easily people forget that civilized society cannot be taken for granted. You have to work for it and renew it always.

[Ramsés V. Morales @rmorale]

@kmele 😅

[Ramsés V. Morales @rmorale]

@lancelotbla, saw you today in Beato's podcast with Mr. Rudess.

Amazing work!

(following your research from now on)

[friendliai @friendliai]

DeepSeek-R1 is now available on Friendli Serverless Endpoints! 🚀
Here’s why teams are excited to integrate DeepSeek-R1 into their workflows:
⚡ Blazing-fast speeds combined with high-quality responses
🔒 Privacy-first hosting in a secure, US/Europe-based data center
📖 Support for maximum context lengths of up to 164k tokens
💰 Flexible, hassle-free pay-as-you-go pricing

Try @deepseek_ai now with a free credit 👉 https://t.co/DozL6a7TWr
Check out our blog 👉 https://t.co/dhZ75K8zdV

#FriendliAI #DeepSeekR1 #ServerlessAPI #LLM #Inference

[Ramsés V. Morales @rmorale]

:-)

[Story @storyandscience]

It's good to be alive! Hello world.