Oriol Pont

stateful and strongly (or eventually) consistent.

[Oriol Pont @uripont_]

good morning

[Oriol Pont @uripont_]

before all of our devices have dedicated inference accelerators, the path forward to bring ~GPT-5 capability and experience to edge devices might be enough with a mix of low-bit weights, KV cache quantization, speculation, and some small variations over the current architectures

[Ahmad @TheAhmadOsman]

I am betting big time on GLM 6

There are many recent papers with great pre-training optimizations (e.g. DSv4)

Now, if Zhipu uses some of that (+ their own novel research), and top it off with their current post-training regime, we're looking at an amazing SOTA in the making

[Oriol Pont @uripont_]

@0xSero nobody has cracked yet what @PrismML did to retrain models to be 1.125bit

[Oriol Pont @uripont_]

@paulg 🤔

[François Chollet @fchollet]

The hardest problems are rarely solved by adding more complexity to the solution -- they are solved by reframing the question until a simpler, clearer answer reveals itself.

[Xenova @xenovacom]

I gave Fable 5 one job: write custom WebGPU kernels for Gemma 4 inference.

It climbed to 84 tok/s, then hit a wall, insisting further optimization was impossible.

Hours later, Anthropic rolled back invisible LLM development safeguards, and it hit 255 tok/s.

The next day, access to Fable 5 was suspended globally.

[Oriol Pont @uripont_]

@initjean rivet mentioned?

[Oriol Pont @uripont_]

@SoloGen @khademinori clear and to the point

[Ray Fernando @RayFernando1337]

For WWDC I was hoping for realtime Siri AI. The UX pattern of holding a button or typing (who types in 2026??) and asking it to do stuff feels like a dying paradigm akin to the floppy disc era of Mac.

I wanted a canvas to take over and go full immersion with voice and personalization. I didn’t want a running chatbot list view on my phone. I wanted the ability to use the SDK to keep user privacy in mind but build unique experiences across the OS.

I want the concept of “apps” to die and for developers to focus on making things magical for users. We are entering a new era, where in the last half of this decade humanity will solve the most complex problems faster than ever. Things should feel like they are part of a bigger system rather than “download my app to fix cancer.”

OpenAI GPT Realtime 2 is incredibly fast and extremely capable right now. I think it’s only a matter of time for people to realize where the puck is going and to push humanity in a new direction for our era of high intelligence.

[Oriol Pont @uripont_]

(and yet no one realized, not even the 1 month discrepancy. someday...)

[Oriol Pont @uripont_]

about 6 months ago

[Oriol Pont @uripont_]

for single-tenant use cases, instead of multiple (different!) experts used for each token generated, which in practice requires all experts to be loaded, it would be nice if we just needed a few expert(s) for a whole response.

[Oriol Pont @uripont_]

@aaronmahlke that this is starting to become real
https://t.co/Nq78iYu1S2

[Oriol Pont @uripont_]

about 6 months ago

[Oriol Pont @uripont_]

@mweinbach sounds familiar...
https://t.co/izx6Cl2Oir

[Oriol Pont @uripont_]

for single-tenant use cases, instead of multiple (different!) experts used for each token generated, which in practice requires all experts to be loaded, it would be nice if we just needed a few expert(s) for a whole response.

[Tim Hua 🇺🇦 @Tim_Hua_]

This graphic looks terrifying kudos to whoever made it.

[Oriol Pont @uripont_]

@NathanFlurry

[Oriol Pont @uripont_]

@ChrissGPT @JasonBotterill is this the way?

[Oriol Pont @uripont_]

@margelo_com we got lynx dual-thread arch in react native in about a year, another win from the margelo team

[Oriol Pont @uripont_]

@ImLunaHey @reactnative hopefully this can be eventually incorporated to Expo Desktop by @birch_js !

[Oriol Pont @uripont_]

@pablocpz_ai @manu__martinm que se vaya pensando recomendaciones locales🗣️