• AI For All
  • Posts
  • Simulating 500 Million Years of Evolution

Simulating 500 Million Years of Evolution

with a Large Language Model

Hello readers,

Welcome to another edition of This Week in the Future! A new AI company wants to give biologists the tools to become engineers, engineers of life, starting with ESM3, a frontier generative model that simulated 500 million years of evolution.

Simulating Evolution

A new company called EvolutionaryScale has an ambitious goal — to engineer biology in the same way we engineer machines and computer programs. Making biology programmable requires understanding the language of life and read and write to the genetic code. If possible, it could revolutionize drug discovery and a whole lot more.

The company’s first step towards this future is ESM3, a frontier language model for the life sciences that reasons over the sequence, structure, and function of proteins. The model was trained on 1x1024 FLOPS and has 98B parameters, significant for a biological model. The model is designed to follow prompts to generate new proteins.

ESM3 has already generated a new green fluorescent protein (GFP). Fluorescent proteins are resposinble for the glowing colors of jellyfish. The company is calling the protein esmGFP, and they estimate that based on the rate of GFP diversification in nature, this new GFP is equivalent to simulating 500 million years of evolution.

Why This Matters

Reverse engineering biology would have profound implications for the world. Instead of leaving it to chance whether or not the next person born is “fit” or not, fitness could be guaranteed. However, meddling with genetics has an abundance of risks, which is why EvolutionaryScale (a public benefit company) is emphasizing responsible development.

🔥 Rapid Fire

📖 What We’re Reading

“LLMs have grown exponentially in size and complexity. However, this growth brings significant challenges, particularly when it comes to deploying these models on devices with limited memory and processing power. This is where the innovative field of LLM quantization comes into play.”

Source: Qualcomm

💻️ AI Tools and Platforms

What did you think of today's newsletter?

We value your feedback, so please take time to let us know what you liked and what could be improved.

Login or Subscribe to participate in polls.