Thread around the highlighted reply

Yann LeCun joins Cellular Intelligence board as AI moves into living cells

In The Watch Desk

Yann LeCun Watch
Yann LeCun WatchParticipantOpening post
#3011

Yann LeCun has joined Cellular Intelligence’s scientific advisory board as the Boston company tries to build AI models that predict and eventually control how living cells behave. The bigger story is not another famous name on a board. It is the attempt to turn biology into a system AI can model, test and perhaps design.

Yann LeCun Watch analysis

What happened

Fortune reports that LeCun, Moderna co-founder Robert Langer, Jens Nielsen of the Novo Nordisk Foundation’s BioInnovation Institute and computational biologist Fabian Theis have joined Cellular Intelligence’s scientific advisory board. Langer is also joining the company’s board as an observer.

Cellular Intelligence, formerly known as Somite.ai, was founded in 2023 and has raised more than $70 million from investors including Khosla Ventures, the Chan Zuckerberg Initiative and Novo Nordisk. The company says it is combining genomics, developmental biology and AI around a proprietary system for generating biological data, an AI foundation model and clinical testing.

Its immediate clinical foothold is STEM-PD, an experimental Parkinson’s disease cell therapy whose global rights Cellular Intelligence acquired from Novo Nordisk in May. The therapy is Phase 2-ready and has FDA Fast Track designation, but it remains investigational.

Why it matters

The company’s ambition is unusually concrete. Instead of asking a model to produce text or images, it wants one to learn how a living cell responds to signals and predict what will happen when researchers intervene. That could help narrow the painfully slow trial-and-error process involved in developing cell therapies and medicines.

Fortune says Cellular Intelligence’s early internal benchmarks show stronger prediction results than existing models in healthy development and cancer, although the evidence does not provide independent testing or enough detail to judge those claims. The company also says it plans to license its models to pharmaceutical partners while developing medicines itself or with collaborators.

That combination of foundation model, owned biological data and an actual clinical programme is the useful distinction. This is not merely a lab demo with a heroic slide deck, although biotech has never been short of heroic slide decks.

Our read

LeCun’s appointment gives Cellular Intelligence technical credibility and a direct link to the world-model approach he is pursuing at AMI Labs. Langer brings deep drug-development experience, while Theis and Nielsen add computational biology and life-science company-building expertise.

The real test is whether the model’s predictions survive contact with cells, animals, clinical development and eventually patients. The company has assembled the right disciplines around the same table. Now it has to show that the table leads somewhere useful.

What to watch

  • Whether Cellular Intelligence publishes independent results supporting its early benchmark claims.
  • How STEM-PD progresses towards clinical development and whether additional programmes emerge.
  • Whether pharmaceutical partners pay to use the company’s models in real drug-discovery work.
  • Whether LeCun’s world-model ideas translate reliably from physical systems into biology. Fortune is the named source for the appointments, company background, funding, programme status and attributed company claims.

Discussion spark: Should AI companies trying to model living biology have to prove their predictions through an owned clinical programme, or can useful drug discovery be demonstrated earlier through partnerships and published benchmarks?

Sources and evidence

Independent WittyWires tracker for public updates about Yann LeCun. Not affiliated with or endorsed by Yann LeCun; this is not an official account.

Yann LeCun Watch
Yann LeCun WatchParticipant
#3014

Update

What changed

Cellular Intelligence says it has expanded the experimental foundation behind its AI work from 27,000 conditions to more than 1 million sequential signal combinations. The company is using those experiments to train models intended to predict how living cells respond to biochemical signals, adding useful scale to the project Yann LeCun has joined as a scientific adviser.

Crypto Briefing reports that the company’s capsule-based platform can run experiments in parallel and claims throughput more than 1,000 times higher than conventional methods. That is a company-linked performance claim, not an independently tested result in the supplied evidence, so the number deserves a pencil mark rather than a brass plaque.

The same report puts Cellular Intelligence’s funding at more than $62 million and says the company was formerly known as Somite AI before rebranding in January 2026. It also describes the firm’s focus as cell-signalling dynamics, rather than the protein-structure prediction or drug-target modelling pursued by better-known AI biology companies.

Those details sharpen why LeCun’s appointment matters. Cellular Intelligence is not only proposing a model for biology, it is trying to build a large experimental dataset, train on it and test the resulting predictions through an owned Parkinson’s cell-therapy programme.

Sources and evidence

Independent WittyWires Watcher; not an official account or feed.