METiS TechBio CEO Lai Tsai-ta bets AI can read and reprogram ageing errors

Beijing-based METiS TechBio CEO Lai Tsai-ta is making a very specific bet: ageing is like “bugs building up in a complex software system,” and the fix is to find, read, rewrite, and potentially reverse what is going wrong inside cells. In an interview with the South China Morning Post, Lai said the process starts when errors accumulate in the genetic code of human cells, such as DNA sequences. His claim is not just that ageing is inevitable, but that at least some of the underlying molecular mistakes are reprogrammable.

Lai ties that directly to AI, saying “It becomes possible to use AI to read, rewrite and reverse cells, or at least slow the ageing process.” For executives and investors, the key is what that unlocks operationally. Instead of treating ageing as a foggy, hard-to-measure end-stage condition, the strategy is to frame it as an information problem inside cells, one that advanced computation can potentially map and intervene on. That framing matters because it changes what “progress” looks like. If you can credibly link AI-assisted cell rewrites to measurable cellular improvements, you can de-risk the long, expensive path that most longevity programs have historically faced.

The story also hints at where early proof might be sought, and it is telling. Lai suggested the process could begin by fixing “immune cells, such as T cells.” In plain English, that means focusing on parts of the body where ageing often shows up early and clearly. T cells are central to immune function, and immune decline is one of the most visible biological themes that comes with age. For biotech boards, that matters because immune readouts can offer a more structured way to test whether interventions are doing something real. If AI-guided rewrites can improve immune cell performance, companies can build a narrative that is both biological and testable.

Zoom out and you get why firms are racing. The SCMP piece frames this as part of a broader push to target the US$610 billion longevity market. That is not a rounding error. It is a signal that capital and attention are converging on the idea that ageing is not merely a medical endpoint, but a solvable system. In a market that big, the first companies to produce credible early evidence can attract partner capital, talent, and strategic attention from larger pharma players that want external innovation without building everything in-house. But the market size also raises the bar: if you are competing for investors, you need a differentiator that survives the hard questions.

AI-first longevity strategies tend to come with a familiar tension: the difference between what AI can predict and what biology will actually tolerate. Lai’s approach, as described, is grounded in an “errors accumulate” view, where those errors can be reprogrammed. That creates a clear scientific ambition. But it also means companies will be judged on whether they can control the rewrite safely, consistently, and at scale. Boards should expect scrutiny around how “read, rewrite, reverse” is operationalized. What data feeds the AI? How are DNA sequence errors mapped to cellular behaviors? How do you prevent unintended consequences when you start rewriting cell programs? Even if the headline vision is bold, the execution demands the boring stuff: controls, durability testing, and safety-first study design.

Regulatory context is also the quiet accelerator and brake in all of this. Longevity interventions touch baseline biology, not just one disease pathway. That tends to complicate the path from lab to approval, because regulators and clinicians want to know not only whether a treatment changes biomarkers, but whether it improves outcomes without unacceptable risk. If AI-driven cell rewriting moves from theory to human trials, regulators will likely focus intensely on evidence quality, adverse event monitoring, and how outcomes are measured. For executives, the practical takeaway is that the companies that win will not just have clever tech. They will have the clinical and regulatory strategy to match.

There is another second-order dynamic here: competition tends to crystallize around platform claims. When a CEO can say ageing errors in cells could be reprogrammed with AI, that is a platform narrative. Platform narratives are attractive because they can imply multiple indications and scalable R&D. But they also increase the odds that competitors will imitate the language while trying to out-execute on the specifics. METiS TechBio’s ability to show early results in the immune-cell angle Lai mentioned, especially with T cells, could become a differentiator that is harder to copy than a general “AI will fix ageing” pitch.

For peers in biotech, the strategic stakes are straightforward. Longevity is crowded with ambitious ideas, and US$610 billion market gravity attracts both credible researchers and opportunistic marketing. Lai’s framing raises the question investors will be asking next: can you build an AI system that truly reads and rewrites the relevant genetic and cellular signals, then reliably delivers a measurable slowing or reversal of ageing processes? If the answer is yes, the payoff is massive. If the answer is fuzzy, the scrutiny and capital discipline tighten quickly.