Alex Laterre

Co-founder @IneffableLabs | prev. Head of Research @ Instadeep

[Alex Laterre @AlexLaterre]

We are launching Ineffable Intelligence with David Silver and my exceptional co-founders!

We’re building a system that can discover all knowledge from its own experience, from elementary motor skills to profound intellectual breakthroughs. We expect it to rediscover, and ultimately transcend, humanity’s greatest inventions: language, science, mathematics, and technology.

We believe this can be built within years.

There is a real risk of failure, in pursuit of a small chance of extraordinary success that could change the course of AI, and with it, humanity.

If this resonates with you, join us. We’re building a world-class team in London.

https://t.co/rcjZtYNBCs

[Ineffable Intelligence @IneffableLabs]

Introducing Ineffable Intelligence. Led by David Silver, we're assembling the best engineers and researchers in the world to make first contact with superintelligence. We’ll be solving the hardest problems in AI on the way. Come join us.

https://t.co/zUuvPJGmcq

[Alex Laterre @AlexLaterre]

@louvishh @VincentMoens I'm in!

[Alex Laterre @AlexLaterre]

Scaling RL is hard. Keeping thousands of actors / learners fed without the hardware fighting you means solving challenges across interconnect, latency, and kernel efficiency.

We are working with NVIDIA to unlock the next scaling paradigm on their latest Vera Rubin platform!

[NVIDIA @nvidia]

We're working with @IneffableLabs to co-design the infrastructure for large-scale, reinforcement-learning agents and accelerate discovery across science and industry.
Our engineers have teamed up to explore how to create the training pipeline that will allow agents to discover breakthroughs across all fields of knowledge.
Learn more: https://t.co/ergHdOIAZF

[Alex Laterre @AlexLaterre]

@alainsamjr @DeepPCB Alumni should receive a life-long subscription. Bring me one next time! ;)

[Alex Laterre @AlexLaterre]

We're hiring exceptional DevOps Engineers to join @ineffablelabs in London.

If you've battled with K8s at scale (1000+ GPUs, hardware failures, autoscaling, storage, scheduler) come build the infrastructure Superintelligence will run on. https://t.co/YcqTYWASpR

[Alex Laterre @AlexLaterre]

@kbeguir @IneffableLabs Thanks K! I will keep carrying forward many of InstaDeep’s original beliefs around RL

[Alfred Lin @Alfred_Lin]

https://t.co/77BHuePAQe

[Alex Laterre @AlexLaterre]

Our new episode of Let’s Talk Research is live 🎙️

Alex Graves unpacks our approach to Generative AI for Biology, driving our progress in proteomics and antibody design with ProtBFN and AbBFN2 🦙

Coming up next: Atomistic Modelling...

[InstaDeep @instadeepai]

We’re back with the BFN story in Episode 2 of the Let’s Talk Research podcast.

Hear Alex Graves dive into real-world applications of Bayesian Flow Networks, including their use in protein sequencing and antibody design. 🧵⬇️

[Alex Laterre @AlexLaterre]

Rust feels built for the era of Agents. Strong typing + strict compiler = the perfect feedback loop LLMs need 🦀

[Alex Laterre @AlexLaterre]

TPUs are fantastic, but most everyday workflows still run on GPUs. Great to see the scaling book broadening its scope; a must-read for our performance engineers!

[Jacob Austin @jacobaustin132]

Today we're putting out an update to the JAX TPU book, this time on GPUs. How do GPUs work, especially compared to TPUs? How are they networked? And how does this affect LLM training? 1/n

[Alex Laterre @AlexLaterre]

@EdanToledo @Meta @ucl Big congrats @EdanToledo! Great to see you putting more works out there. I might even give this one a read 😱😉

[Alex Laterre @AlexLaterre]

@tw_killian @instadeepai Thanks Taylor! Indeed big congratulations to the whole team @instadeepai for these amazing achievements 💪

[Alex Laterre @AlexLaterre]

Going from our first Nature journal paper in late 2024 to four published in the first half of 2025, is mind-blowing! 🤯

[InstaDeep @instadeepai]

📆 Six months, four publications, one cover.

We’re halfway through the year and InstaDeep research is powering ahead with multiple boundary-pushing papers published in the @NaturePortfolio!

Catch up on the highlights below 🔽

[Alex Laterre @AlexLaterre]

We made the cover of @NatMachIntell! 🌱

ChatNT is a Conversational Agent analysing genomics sequences to answer key biological questions, assisting scientists in their work 👩‍🔬

Kudos to @deAlmeida_BP, @thomas_pierrot & the @instadeepai research team for this huge milestone!✨

[Machine Learning for Protein Engineering Seminar @ml4proteins]

Next Tues (6/17) at 4PM ET, we will have
Bora Guloglu and @tomdbarrett present "Flexible Antibody Design with Multimodal Bayesian Flow Networks"

Paper: https://t.co/pX5e4hIxwb

This will be followed by @BrianHie, who will give an early-career talk on Wed (6/18) at 4PM ET!

[InstaDeep @instadeepai]

Pairing speed 🪽with near-quantum accuracy 🔍—experience both in atomic and molecular behaviour models with mlip, our open-source library for working with Machine Learning Interatomic Potentials (MLIP).

🧵Read the thread below to learn more.

[Alex Laterre @AlexLaterre]

Our latest antibody foundation model, AbBFN2, is now live as an experimental workflow on https://t.co/mgynHW9o0c 🌱

It’s one thing to contribute to the scientific community by publishing. It’s another to see your models being deployed. I'm excited to team up with industry partners to explore practical applications of this new model 💥🧬

[Alex Laterre @AlexLaterre]

Joining InstaDeep has its perks -- and surprises. Just ask Bora, one of our AI Scientists, who found himself on stage introducing AbBFN2, our new foundation model for antibody design 🎙️

... and yes, he nailed it 💪

[Alex Laterre @AlexLaterre]

Huge congrats to the brilliant @InstaDeepAI team behind AbBFN2! 🎉 Paper & model are now live on @huggingface and code on GitHub. Give it a try!

📖 paper: https://t.co/spaHna1aHO
🤗 Model: https://t.co/zBPb3aGT5C
⭐ Code: https://t.co/Vlr70exS4N

[Alex Laterre @AlexLaterre]

Just weeks after releasing ProtBFN/AbBFN in @NatureComms, we're back with AbBFN2 ⚡️

Our new Antibody Foundation Model goes beyond sequence, modeling 45+ data modalities incl. genetic & biophysical properties. This creates a rich grammar for guiding antibody design.

[Alex Laterre @AlexLaterre]

AbBFN2 translates our mission into action: modeling the joint distribution of scientific metadata across modalities -- laying the foundation for a generative, holistic view of biology.