Background

On March 10th, 2026, DayOne announced that it will build biological data centers in Singapore and Melbourne. "Biological" here means that the computational infrastructure will involve live human neurons grown on silicon chips, maintained by life support systems that keep the cells alive through nutrient delivery and waste removal. Human brain cells are efficient information processors, and it is not surprising that commercial interests have begun to take notice. Cortical Labs, the "Wetware-as-a-Service" company discussed in prior posts (here and here), is the company supplying the CL1 biocomputers. DayOne's Singapore facility has the potential to scale up over time to house 1,000 units of CL1, representing a deal that could grow to be worth tens of millions of dollars.^[1]

DayOne is one of the fastest-growing data center operators in the world, with plans to raise a $5 billion IPO in the United States later this year at a targeted $20 billion valuation. By shifting workloads from pure silicon to increasingly sophisticated configurations of human brain matter, companies like DayOne may be able to offer cloud computing services in a more energy-efficient manner. They can frame these operations as sustainability initiatives, given that, as they state in their official announcement, a "bio data center harnesses the natural efficiency of brain-like organoids, which can function on a fraction of the wattage required by digital computers."^[2] Interestingly, opponents of placing human neurons in server racks to power the next generation of AI, because of the current regulatory vacuum and the worrisome incentives this creates, can thus be framed as opponents of ESG and sustainable computing. 

It is entirely uncertain how quickly biological computing will advance, or how commercially viable DayOne's bio data centers will ultimately prove. The current generation of the CL1 recently went viral for playing the video game Doom^[3] and for being connected to an LLM.^[4] These demonstrations may sound impressive or underwhelming depending on your familiarity with the field, but it is worth noting that deep learning systems in the 2010s scaled from a niche research approach to the dominant ML paradigm. What is certain is that biological computing is not currently just being evaluated in carefully regulated research contexts. Instead, it is being pushed at commercial scale into mainstream data center operations, in a regulatory vacuum by well capitalized companies with strong incentives to downplay risks.

The Problem

A revenue model is established, institutional capital is committed, and one of the fastest-growing data center operations in the world has publicly committed to commercialized biocomputing. It is certainly the case that current biocomputers are rudimentary, and my belief is that current systems are not sentience candidates. But every dollar invested in biological data centers is a dollar that depends on the continued assumption that these systems do not matter morally. The commercial pressure to build increasingly complex and capable systems to handle inference workloads means that this assumption will be stress-tested at exactly the moment it is most expensive to abandon. The acute risks of an entrenched "brain farming" industry, in which novel forms of consciousness are exploited at scale for computational purposes, are no longer hypothetical.

Proposal

The request is simple: an immediate moratorium on the use of biological neural tissue as commercial computing infrastructure.

Next Steps

Companies should be restricted from building data centers that leverage the computational power of biological systems, and further development of these systems should occur in regulated research environments without ties to widespread commercialization.

Anyone reading this should first and foremost understand that the focus here is not on research, but rather on the widespread commercialization of such systems before adequate safeguards have been addressed. The below are ways in which I believe we can effectively work to prevent the formation of a commercial ecosystem that will be extremely difficult to challenge once entrenched:

• Journalists covering this space should convey not just the technical novelty of biocomputing, but also the governance gap in which the technology is being commercialized and the incentive structure that early commercialization creates.
• Regulators should be asked, formally and on the record, whether their mandates cover the commercial deployment of human neural tissue as computing infrastructure.
• Donors who provided biological samples from which these neurons are derived should be informed that their cells may be used as commercial computing infrastructure in DayOne's data centers.
• Institutional investors of DayOne, and the banks underwriting the IPO (JP Morgan and Morgan Stanley), should be briefed on the ethical and regulatory risks of their proposed biological data center buildout.
  • It is highly unlikely, in my opinion, that the investors backing the company have been accurately made aware of the situation. Religious and political interest groups are likely to rally against DayOne and the commodification of human neural tissue in the near future, and this decision by DayOne was probably made after investor capital was already secured and without institutional endorsement.
• The scientific community should establish a coordinated ethical response that emphasizes that research and widespread commercialization are distinct concepts and that the latter should not proceed in the absence of regulatory frameworks.

The DayOne announcement has generated widespread media coverage, but there has been no serious pushback to the worrying incentives that such early commercialization creates. There has been no strategic action taken to address the risks involved. This needs to change.

Please feel free to reach out if you're interested in getting involved. Although I think that anyone who finds the above arguments compelling should feel empowered to act on them individually (with the aforementioned framing). Happy to discuss alternative solutions or opposing viewpoints.

1. ^{^}

   Cortical's CL1s have a $35,000 list price. After the validation phase, the Singapore facility may expand to up to 1,000 units (implied contract value of $35M, but there are likely steep volume discounts).

2. ^{^}

   https://dayonedc.com/sustainability/dayone-and-cortical-labs-to-develop-singapores-first-biological-data-center

3. ^{^}

   https://www.newscientist.com/article/2517389-human-brain-cells-on-a-chip-learned-to-play-doom-in-a-week/

^{^}

https://x.com/dr1337/status/2028619242275475463?s=20

Show all footnotes