Silicon-Based Intelligence vs. synthetic biological intelligence: what is the difference?

What is synthetic biological intelligence?

Synthetic biological intelligence is computing with living neurons grown on a chip. The term comes from a 2022 paper in the journal Neuron by Brett Kagan and colleagues at Cortical Labs, a company based in Melbourne (Wikipedia, "Cortical Labs"). They called their system DishBrain.

The paper starts from a motivation, not a claim about AI: Integrating neurons into digital systems may enable performance infeasible with silicon alone.^↗ The team grew networks of human and rodent neurons on a high-density multielectrode array, then connected them to a simulated version of the arcade game Pong (Kagan et al., 2022). Electrodes relayed the position of the ball and paddle to the cells. A miss triggered random stimulation and a hit triggered ordered stimulation (Wikipedia, "Cortical Labs"). With this closed-loop feedback, the cultures showed apparent learning within five minutes of play, and the authors named what the cells were doing:

“which we term synthetic biological intelligence”

The BBC reported that the culture held about 800,000 brain cells, some grown from human stem cells and some from mouse embryos, and that it often missed the ball (BBC News, 2022).

The CL1

In March 2025 Cortical Labs launched the CL1, which it calls the world’s first commercial biological computer (DCD, 2025). It was announced at a price of $35,000, with about 800,000 human neurons (Wikipedia, "Cortical Labs"). The company describes how it works:

“Real neurons are cultivated inside a nutrient rich solution, supplying them with everything they need to be healthy. They grow across a silicon chip, which sends and receives electrical impulses into the neural structure.”

The company presents its Synthetic Biological Intelligence as a more advanced and sustainable form of AI^↗ As of October 2026, Cortical Labs still offers the CL1 for sale on its website (Cortical Labs, CL1).

What is organoid intelligence?

Organoid intelligence is the closest related term. In February 2023, a team led by Lena Smirnova and Thomas Hartung of Johns Hopkins University published a paper in Frontiers in Science on human stem cell-derived brain organoids, which are 3D cultures of human brain cells grown in the lab. They wrote:

“Coining the term “organoid intelligence” (OI) to encompass these developments”

Brett Kagan of Cortical Labs is one of the paper’s co-authors (Smirnova, Hartung et al., 2023). The two terms overlap: both describe using living neurons for computing. Synthetic biological intelligence names what the DishBrain cultures did. Organoid intelligence names a wider research program built on brain organoids.

What is Silicon-Based Intelligence?

Silicon-Based Intelligence is this site’s term for AI that runs on silicon chips: language models, image models and the other systems that learn from data. A large language model, for example, is a neural network trained on a vast amount of text (Wikipedia, "Large language model"). All of its computing happens in transistors. No living cells are involved. The full definition is on What is Silicon-Based Intelligence?

How do they differ?

Both approaches use silicon chips, which is part of why the names are easy to confuse. The difference is what does the computing.

	Synthetic biological intelligence	Silicon-Based Intelligence
What computes	Living neurons, grown from stem cells	Silicon transistors
What the chip does	Stimulates and records the neurons	Everything: it is the computer
How it learns	The cells change their activity in response to feedback, as in DishBrain’s Pong game	The model is trained on large amounts of data, such as text
Scale	About 800,000 neurons in DishBrain and the CL1	Clusters of chips; see the energy figures below
Kept alive or running by	Nutrients from a life-support system	Electricity

On scale, the organoid intelligence paper gives a useful frame. It estimates that a human brain has 86 to 100 billion neurons and runs on about 20 watts. In the same paper, the Frontier supercomputer drew 21 megawatts, and the clusters used to train state-of-the-art machine learning models typically ran at around a million watts (Smirnova, Hartung et al., 2023). The authors add that brains and machines perform quite different tasks. A culture of 800,000 neurons is a tiny fraction of a brain, and these figures date from 2023.

What ethics questions does each raise?

Each side raises its own questions, and neither has settled answers.

Living neurons. The DishBrain paper’s title describes the neurons as showing sentience. In the paper, sentient refers to how the cells react to their environment, not to feelings or sensations (Wikipedia, "Cortical Labs"). The BBC summed up other experts’ view as calling the brain cells sentient is going too far^↗, and added that with no consciousness, it does not know it is playing Pong in the way a human player would, the researchers stress^↗ (BBC News, 2022). The organoid intelligence authors expect harder questions as the systems grow more complex:

“Could organoids experience pain and, if so, would they suffer – even in rudimentary ways?”

They propose an embedded ethics approach, in which ethicists, researchers and members of the public review the work as it develops. Privacy for the people who donate the cells is among the other issues they list (Smirnova, Hartung et al., 2023).

Silicon chips. For AI, the question is whether software running on transistors could ever have experiences. A 2023 review by consciousness researchers found that no current AI systems are conscious, but also suggests that there are no obvious technical barriers to building AI systems which satisfy these indicators^↗ (Butlin et al., 2023). Anthropic, which builds AI models, says There's no scientific consensus on whether current or future AI systems could be conscious^↗ (Anthropic, 2025). Others, including Microsoft AI chief Mustafa Suleyman, argue that We should build AI for people; not to be a person.^↗ (Suleyman, 2025). The wider debate is on Is AI conscious?

This page reports both debates. It does not claim that either kind of system is alive or conscious.

Why the abbreviation clashes

Synthetic biological intelligence and Silicon-Based Intelligence share the same initials. Cortical Labs uses the abbreviation for its own work. Launch coverage of the CL1 describes a neural network dubbed Synthetic Biological Intelligence (SBI)^↗ (DCD, 2025). The same letters also stand for State Bank of India.

The two terms point in opposite directions. Synthetic biological intelligence is carbon-based: real brain cells do the computing. Silicon-Based Intelligence is computing done by silicon itself. To keep them apart, this site always writes Silicon-Based Intelligence and synthetic biological intelligence in full. The other names in this debate are in the glossary.