Daniel Hernández

[ℏεsam @Hesamation]

this 3b1b visualization of how LLMs work layer by layer is the true brain rot we never knew we needed.

[Daniel Hernández @danihv10]

Este año se están sentando las bases de las aplicaciones basadas en modelos fundacionales, y no me gustaría que proyectos basados en trabajos de una comunidad que fomenta la cooperación, intenten crecer de forma desleal en vez de encontrar su propia voz en el mercado.

[Daniel Hernández @danihv10]

Como Murciano y alguien que se pelea continuamente con modelos generativos, este post de @javilop me ha tocado la patata 🫶 ⤵️

[Javi Lopez ⛩️ @javilopen]

Some reflections I'd like to share with you :)

I suppose when the competition uses our name (Magnific) for their clickbaits or marketing campaigns, I should feel flattered.

It's been insane, the attention we've received these past months, with everyone: Leonardo, Krea, Freepik, Scenario, Playground... Even Stable Diffusion, and loads of Reddit threads with ComfyUI/Automatic1111, trying to launch their own upscalers and using us as a benchmark.

It's mind-blowing that two nobodies from Murcia (a very small city in Spain) have been the center of this attention. All of this seems like absolute madness to both Emilio and me!

There was a gap in the market that Magnific filled, and nobody had realized it until then.

And now, it has a name. A name that all of us who experienced the first weeks of Magnific's journey coined together with affection:

REIMAGINED UPSCALING.

And it's normal for other competitors to want to be there! Not only is it normal, it's also good for everyone!

For you, because you have more options available, and for Emilio and me, because it pushes us to up our game!

Though honestly, I feel sorry when some try to deceive people by making them believe they have "the one Magnific formula" when in reality, they don't. Because the Magnific formula consists of a lot of small pieces and fine tuning that are impossible to replicate.

But more importantly... THEY DON'T NEED "the Magnific formula"! Because you know what? Each upscaler could be unique, after all, and will always have its own touch!

Will there be better upscalers than Magnific in the future? Of course! But for some, the best will always be Magnific, because we will always have our own mojo, our own style. Yes, even in something like upscaling. And this, of course, applies to other competitors as well. And hey, we won't get jealous :)

So really, we are honored to be considered the benchmark, but... And in this, unusual for me, I want to get serious, there's no need to try to deceive people by saying "we are the new Magnific" or "we have the Magnific formula".... Because you know what? Not only is it untrue, and therefore you're lying, it's mainly because you don't need that to be a good upscaler; you just need your own touch!

Competitors, let's play fair, please! There is cake for everyone. Don't deceive people. Be fair. And be honest.

Cheers!
Emilio and Javi.

[Daniel Hernández @danihv10]

Si en vez de eso, el objetivo se convierte en suplantar, sin ánimo de ofrecer un ápice de personalidad, todo lo bueno que podemos esperar de la competencia se perderá.

[karthik @akbirthko]

something about this aesthetic is arousing

is it the computer modern sans? the shapes?

monke see monke happ: https://t.co/rkwxZC1AwF

[Jack Morris @jxmnop]

today i found out that this one australian guy has been toiling away making incredibly detailed Neural Circuit Diagrams with the vibe of a 1950s issue of Popular Mechanics, but content fit for the 2020s

behold. the Transformer

[Daniel Hernández @danihv10]

@mrstartups 🙌

[Daniel Hernández @danihv10]

Empieza una nueva epoch 👀

[Sam Altman @sama]

here is GPT-4, our most capable and aligned model yet. it is available today in our API (with a waitlist) and in ChatGPT+.

https://t.co/2ZFC36xqAJ

it is still flawed, still limited, and it still seems more impressive on first use than it does after you spend more time with it.

[Daniel Hernández @danihv10]

@nacmartin Que gran utilidad!!

[Daniel Hernández @danihv10]

@tonilopezmr 🙋‍♂️

[Daniel Hernández @danihv10]

@javilop @DotCSV @pau

[Sciling @scilingai]

Mira como la #InteligenciaArtificial convierte #despechada de @rosalia en un cuento de #Navidad. Entra en https://t.co/eUwPOBp3vi y crea tu postal inteligente.

[Sciling @scilingai]

Mira como la #InteligenciaArtificial convierte #despechada de @rosalia en un cuento de #Navidad. Entra en https://t.co/eUwPOBp3vi y crea tu postal inteligente.

[UCAM Murcia @UCAMMurcia]

/𝑪𝑶𝑵𝑪𝑼𝑹𝑺𝑶 #𝑫𝒆𝒔𝒅𝒆𝑬𝒍𝑪𝒆𝒏𝒕𝒓𝒐/

⛽ ¿Eres capaz de enchufarla?

🚘 Podrás ganar 𝟯𝟬𝟬€ en gasolina para gastar en V2 Gasolineras.

🔂 ¡Haz RT!

✅ ¡Sigue a @UCAMMurcia y @ESTHADERV2!

🎟️ Sorteamos 1 entrada doble entre todos los participantes

🦸‍♂️ #TiempoDeHéroes

[Daniel Hernández @danihv10]

Un recorrido visual por la historia de las redes neuronales

[Philipp Schmitt @philippschmitt]

New research-y project: Blueprints for Intelligence, a visual history of artificial neural networks from 1943 to 2020
https://t.co/CcPFYIxeJz