It’s alive! AI runs on real human cells for first time

In what could well be a major step toward truly ‘alive’ artificial intelligence (AI), scientists have created the world’s first ‘biological computer’ that fuses human brain cells with silicon hardware to form fluid neural networks and which will reach users in the coming months.

As it happens, Australian company Cortical Labs has just launched the CL1, also known as a Synthetic Biological Intelligence (SBI), in Barcelona, Spain, as a more dynamic, sustainable, and energy-efficient model than any AI currently in existence, according to a report published on March 3.

Per the report, the scientists expect their CL1 system to become a game-changer for science and medical research, thanks to the human-cell neural networks that form on the silicon ‘chip’ and present an ever-evolving, ever-learning organic computer.

Indeed, the CL1 model learns so quickly and flexibly that it entirely surpasses the silicon-based AI chips deployed in training existing large language models (LLMs) like ChatGPT and Gemini. 

How the CL1 model works

As its developers explained, it involves cultivating real neurons “inside a nutrient-rich solution, supplying them with everything they need to be healthy.” Then, they “grow across a silicon chip, which sends and receives electrical impulses into the neural structure.”

Specifically, it all takes place in the lab’s proprietary Biological Intelligence Operating System (biOS), running a simulated world and sending straight to the neurons the information about their environment. In response, neurons produce impulses that affect their simulated world.

This way, the scientists bring the neurons to life and integrate them into the biOS using a combination of hard silicon and soft tissue, facilitating direct connection to the neurons, deploying code to the real neurons, “and solving today’s most difficult challenges.”

Overall, each CL1 unit functions independently and features a completely programmable bi-directional stimulation interface, life support and recording systems, and a Python API, which allows for dynamic applications. It is also energy-conserving, with one rack of 30 CL1s requiring only 850-1,000 watts.

In the words of Cortical founder and CEO Dr Hon Weng Chong, the recent CL1 launch is the product of a series of critical breakthroughs and “culmination of a vision that has powered Cortical Labs for almost six years,” adding that customers would be able to buy the CL1 model biocomputer outright or access the chips remotely via the cloud:

“We’re offering ‘Wetware-as-a-Service’ (WaaS). (…) This platform will enable the millions of researchers, innovators, and big-thinkers around the world to turn the CL1’s potential into tangible, real-world impact. We’ll provide the platform and support for them to invest in R&D and drive new breakthroughs and research.”

Elsewhere, advances in other areas of AI technology continue unabated, with the launch of OpenAI’s Sora and ChatGPT 4.5, xAI’s Grok 3, and Anthropic’s Claude 3.7 Sonnet, as well as the proliferation of Robotaxis across the world, and the use of AI on the frontlines and in medicine.