JuanToño

Identity doesn’t fit in a tagline bio… If it does, it probably isn’t very interesting.

[JuanToño @juantono42]

This website for tracking year progress is simple and nice... already at 0.22% at the moment of writing this
https://t.co/arsO07291f

[Trevor Blackwell @tlbtlbtlb]

Remember when compilers would detect that someone was using it to build another compiler and silently inject bugs?

[Dog Years DAO @DogYearsDAO]

𝙒𝙚𝙡𝙘𝙤𝙢𝙚 𝙏𝙤 𝙅𝙪𝙣𝙚 🐾

Another month of building something we genuinely believe in.

This month we're continuing to share the science behind canine aging, what we're learning, and what we're working toward.

If you're new here, we're currently working towards advancing canine longevity through community-owned science and our upcoming clinical trial evaluating follistatin gene therapy in dogs diagnosed with osteoarthritis.

We're incredibly grateful to our community of researchers, veterinarians, dog parents, and contributors who make this mission possible through their support and dedication.

As we step into a new month, we'd love to hear from you

What topic about canine aging, longevity, or dog health would you like us to explore this June?
Drop the in the comment section.

Want to follow our journey more closely this month?

Join our Discord community for updates, discussions, and behind the scenes insights as we continue working toward advancing canine longevity through community-owned science.

https://t.co/xwxhGUr8lc

[BijanBowen @bijanbowen]

Had Claude Fable 5 log network packets and display them as cars on a highway, different car types = different packet types

[Matias Scalbi @MatiasScalbi]

🔥 Esto es una locura, seguí el paper de ayer y me puse a mirar acción por acción en Estados Unidos

El mismo activo, comprado solo de noche contra solo de día, durante años:

- AMD: de noche +4.555.517%, de día menos 99,94%
- Micron: de noche +16.170.077%, de día menos 99,95%
- Nvidia: de noche +221.715%, de día menos 51%

El que operó el intradía, entrando y saliendo para cazar cada movimiento, se fundió. El que compró y se quedó quieto, se llevó todo

Tiempo en el mercado > no timing del mercado

También me da para pensar, capaz crear una estrategia en Hyperliquid para cazar los movimientos nocturnos y hedgear en la apertura no es mala. Si lo hago, comparto luego resultados

[David @dzhng]

don't use loops, design state machines

[Ben Dicken @BenjDicken]

Ya'll laughed at me for loop benchmarks.

Now it's the future of coding.

Hope Boris is using C and Rust exclusively.

[Miles Cranmer @MilesCranmer]

This is an insane paper and I love it

https://t.co/DP8OR5NJf2

[The Kobeissi Letter @KobeissiLetter]

What just happened?

The S&P 500 just erased nearly -$2 TRILLION of market cap just hours after 3rd strongest US jobs report in 18 months.

Meanwhile, Bitcoin is officially down over -50% from its record high in October 2025.

What's happening? Let us explain.

(a thread)

[JuanToño @juantono42]

X allowing video replies is a terrible idea. Many 'influencers' will post minute long videos over explaining and milking other posts core ideas for engagement when they could simply compress and put clearly their idea into a text post. @nikitabier @elonmusk

[fofr @fofrAI]

Fun new feature landing in Nano Banana 2 today:

video-to-image

Pass a video or YouTube URL to Nano Banana 2 along with your prompt, and it'll use the video as context to make the image.

> create a comic strip from this video

[Okan Can @0kncn]

Seedance 2.0 ile yeni AI video denemesi.

Küçük bir elektrik yaratığı, dev kristal boss’a karşı arenada savaşıyor.

Önce GPT Image 2 ile 9 shot’luk storyboard oluşturdum, sonra aynı aksiyon akışını Seedance 2.0 ile videoya çevirdim.

Kullandığım video prompt aşağıda 👇

[Max Zhaoshuo Li 李赵硕 @mli0603]

This is THE moment of Physical AI!

We are officially announcing Cosmos 3: Omnimodal World Models for Physical AI 🚀

- Cosmos 3 is an omnimodal world model: within a unified architecture, it can understand and generate language, images, video, audio, and actions.
- It is not just a VLM, not just a video generator, not just an audio-visual generative model, and not just a physics simulator / world-action model. It can understand images and videos, generate images, videos, and audio, simulate future worlds, predict actions, and generate robot policies—enabling models to truly begin to “touch the world.”
- Cosmos 3 is the #1 open-weight reasoner / T2I / I2V / robot policy across many benchmarks.

Huge thanks to every teammate who fought side by side on this journey—from architecture, data, training, infra, serving, and evaluation to post-training. Every part of this project carries an incredible amount of hard work. This was my first time leading a project as Tech Lead, and I feel truly fortunate.

The future of Physical AI needs models that can not only “see” and “describe” the world, but also “imagine,” “simulate,” and “act”—and eventually close the loop with the real world. I hope Cosmos 3 can become an important starting point for this direction, and I’m excited to push Physical AI into its next stage together with the open-source community.

Welcome to the era of Physical AI.

HuggingFace: https://t.co/QW5h5pIWWM
Project Website: https://t.co/Jppa0gkn16
Code: https://t.co/aJgaLm5BaG

[Chris Paxton @chris_j_paxton]

This is actually really cool

[Happycapy @happycapyai]

Okay, now everyone needs to learn Happycapy aerodynamics.😂

[JuanToño @juantono42]

The industry uses a hybrid approach: we don't write manual mathematical proofs for everything, but use automated mathematical guardrails in our craft tools. Dijkstra was wrong about how the horizontal workflow/execution, but he was right about the process vertical reality…

[Robin Hanson @robinhanson]

"Dijkstra said … Programming is not a craft. It is closer to mathematics than to carpentry, and the moment you treat it as a craft, you guarantee that the software you produce will be full of the kind of bugs that craftsmanship cannot catch. The fix, in his view, was to teach programming the way mathematics is taught. You should be able to prove your program correct before you run it."

Don't we have a half century of experience showing he was just wrong?

[The OSINT Newsletter @osintnewsletter]

📰 An advanced 2026 guide to investigating a person of interest, covering tools, tactics and methodology.

Structured enough that you can lift it into a Claude skill and add your own tradecraft on top.

https://t.co/BStPHDwlLR

[JuanToño @juantono42]

@dwarkesh_sp Karpathy is a nice guy but come on… the argument of “Over the last 200 years, we've automated away a lot of hard physical labour. But people still go to the gym” is simple stupid… multiple factors model the act of going to the gym or abstracting more the idea… doing exercise.

[Startup Archive @StartupArchive_]

Ben Horowitz: “Do not expect life to be fair. It will only defeat you.”

The Andreessen Horowitz co-founder is asked for advice that he has found useful in his own life. Ben responds:

“The thing that I would say has had the biggest effect on me is something my father said to me years ago: ‘Life isn’t fair.’ That advice seems really simple, but the thing I’ve seen that defeats people more than any other thing in life is the expectation of some fairness.”

He continues:

“There are all kinds of things that are going to happen to you that don’t happen to other people and are completely unfair. But it doesn’t matter because that’s the way it is. And as soon as you can get that idea out of your mind [that life should be fair], you can just deal with it… ‘What should I do now?’ is the real question — not ‘How do I go back and get people to be fair?’… Life’s not fair. That’s the nature of it. If you think about it more than five seconds, you’ll realize that… As an individual, do not expect anything to be fair. It will only defeat you.”

Source: @lennysan (Sep 2025)

[Y Combinator @ycombinator]

"You need to make yourself a big target for luck, and the way to do that is to be curious. Try lots of things, meet lots of people, read lots of books, ask lots of questions."

— Paul Graham, How to Do Great Work

https://t.co/AyIk5fnKZL

[Jimmy Heaters @CathPoaster]

new grads often ask me what they should be doing so they don't fall behind in the ai space. there's a lot, but its honestly super manageable. become intimate with model internals. proof based linear algebra. non-convex optimization. this is stuff you could've done in undergrad. it definitely takes some time and work, but its doable. have taste, have opinions. train a small model, then train a big one. vLLM internals, tensor parallelism. hand roll kernels. cluster orchestration. do you have opinions on synthetic data? why don't you? SFT, PPO, you should know this. learn Triton. everyone is reproducing papers now so you need to be doing more. do you know the semi supply chain? where are the bottlenecks? hardware, man, hardware. your little gpu rig erector set in your basement isnt gonna cut it. build a cluster, a big one. pretrain a 800B model. now postrain it. serve it to millions of people. you should be able to beat deepseek on some benchmarks now. its a lot to take in but it all snowballs. this what job security looks like from now on. do you want to work in tech or not