Max Fällström

democratizing robotics

[Qualia @QualiaRobotics]

Foundation models for robots are coming fast.

Qualia is the bridge. We bring foundation models into real factories.

We're hiring a Research Engineer to join our Policy team. VLAs, RL, world models. Real robots, real impact.

Apply: https://t.co/Mzphtg1QzI

[Max Fällström @maxfallstrom]

@GeneralistAI What’s the throughput per hour?

[Qualia @QualiaRobotics]

Qualia will be at @iclr_conf this spring!

This year, the International Conference on Learning Representations has more robotic foundation model research submitted than ever in the conference's history - the field is evolving very rapidly and we are more excited than ever.

We'll be there connecting with researchers, engineers, and anyone building at the intersection of learning and robotics.

Find Ludvig, Max, and Quantao

[Qualia @QualiaRobotics]

Qualia is coming to Shenzhen China 🇨🇳

We are partnering with MakerMods, AMD, and others for a robotics hackathon in one of the global hearts of physical AI.

This weekend hundreds of engineers will be building projects on hardware, software, and models built for a new type of agent.

We look forward to sharing more in the coming days 🤖

[Qualia @QualiaRobotics]

We are hosting an agent meetup in Copenhagen https://t.co/i0q5r36hJh ​
Builders sharing how they actually use AI agents at work. No pitches, no product demos. Just real setups and honest takes. Qualia office in Copenhagen. Drinks. Pizza. Good people. Yes there are robots here too. In person only. Sign up via the link

[Qualia @QualiaRobotics]

The teams that will deploy the most robots in the next 5 years aren't foundation model researchers. They're domain experts in logistics, agriculture, manufacturing, and more. They need infrastructure, not more papers.

Companies will be started all over the world because the models are becoming more capable - we want to be part of your journey from day 1.

Check out what we are building at https://t.co/DT2srK5SyK to solve this.

[Qualia @QualiaRobotics]

Qualia is heading to San Jose for NVIDIA GTC 2026! 🤖

@Fabian13Kerj and Qualia are touching down in San Jose for the conference. We are looking forward to catching up with our users, friends, and meeting fellow builders in the robotics space.

Want to grab a coffee or chat?
👉 Drop us a message here on X.
👉 Or reach out to Fabian directly.

Let's make it happen!
#GTC26

[Shane Legg @ShaneLegg]

AGI is now on the horizon and it will deeply transform many things, including the economy.

I'm currently looking to hire a Senior Economist, reporting directly to me, to lead a small team investigating post-AGI economics.

Job spec and application here: https://t.co/VAfwrMc8Tp

[Max Fällström @maxfallstrom]

If you write syntax code in 2026, you're gonna get left behind

[Ryan Dahl @rough__sea]

This has been said a thousand times before, but allow me to add my own voice: the era of humans writing code is over. Disturbing for those of us who identify as SWEs, but no less true. That's not to say SWEs don't have work to do, but writing syntax directly is not it.

[Qualia @QualiaRobotics]

Some words from one of our founders. Qualia is opening up to more on the waitlist!🤖

[Qualia @QualiaRobotics]

Have a great weekend everyone. Big week coming for Qualia - JAN 15TH

[Qualia @QualiaRobotics]

Lets bring foundation models for robots out there!

[Qualia @QualiaRobotics]

Did you know both Gr00t and xVLA allow you to specify your embodiment?

Sneak peek: Our engine handles embodiment specs across VLAs seamlessly. No more dependency limitations. Coming in Phase 1 for January 15th! Join waitlist: [https://t.co/ZTbj0goqpJ] #FoundationModels

[Fabian Kerj @Fabian13Kerj]

Incredibly excited to introduce @QualiaRobotics to the world 🤖

The long tail of robotic applications is massive, but unlocking it requires great software. We saw this with LLMs, and now we are seeing it with robotics.

At Qualia we are building the first VLA-native platform.

As foundation models evolve, Qualia provides the infrastructure to make your data VLA-trainable across different models, scenarios, horizons and more. We are working towards becoming the development layer for the next generation of spatial intelligence.

Check us out and join the waitlist: https://t.co/XLr2zK3s4r

Thank you to my incredible team for being part of this journey: Ludvig, Max, Oscar, Apoorva, Matan, let begin!

And to our investors for supporting us! Shoutout to
Heartcore Capital, Florent Venture Partners, Wave Ventures, and others ♥️

Next up, opening up Qualia to more people - January 15th!

Happy to be doing this in the Nordics and in Europe.

[Max Fällström @maxfallstrom]

@qwertyu_alex Looks awesome dude! Would be even better with another background and a focus effect on you I think🚀

[Qualia @QualiaRobotics]

Turn your keyboard into a robotics lab with Qualia. Our waitlist is growing and we'll prioritize early signups. Thanks for the great responses so far. We are very excited. https://t.co/ZTbj0gnSAb

[Max Fällström @maxfallstrom]

RT @QualiaRobotics: Excited that Qualia is partnering with NVIDIA and the inception program.

NVIDIA has some incredible robotics tools - w…

[Fabian Kerj @Fabian13Kerj]

@physical_int killing it as again. Vision-language-action models for robotics has been on a rapid improvement rate for the last months - but one of the core problems has been that action chunks produced in isolated sequences causes uncomfortably shaky robots during inference. This latest paper attacks just that: by essentially loading the action state and updating it by responding to the encounters during the previous chunk's execution - the output becomes smooth af.

If you think about it - this is much more akin to how biological brains work and feels like a no-brainer. But to create a solution in practice is an incredible advance. Given that the Pi VLA family outputs actions similarly to diffusion models - this new method is akin to 'inpainting' for those who have played around with image diffusion models.

Only sad thing is that this work is largely done on top of the Pi 0.5 model - which isnt opensource yet - @physical_int amazing job! When can we use your latest model in @QualiaRobotics?

[Fabian Kerj @Fabian13Kerj]

The future of robotics isn't better hardware. It's making it usable for everyone.

True progress is when the biologist, the small business owner, or the problem-solving kid can build a spatial agent without writing a single line of code.

We're making that a reality at Qualia.

[Max Fällström @maxfallstrom]

#tbt Vigårdas vita dipsås, ni vet vilken jag pratar om. Går den att köpa någonstans? Finns det recept?