Dom says AI has turned intelligence into a tool as synthetic biology advances

Dominic Williams has weighed in on one of the most closely watched developments in artificial intelligence and biotechnology, arguing that humanity is entering a period unlike any before.

Reacting to reports that an AI model helped design viruses that had never existed in nature, the Internet Computer founder wrote on X: “After 20,000 generations of humans, this is the first where: intelligence is a tool; we have manufactured ‘life.’ Meanwhile, more than half of the humans that have ever lived, are alive today, and the climate is changing. Interesting times…”

His post followed a New York Times report on research led by scientists at Stanford University and the Arc Institute, which was published in the journal Science. The study described how a genomic AI model called Evo was trained on trillions of DNA building blocks, known as nucleotides, and then used to generate new viral genome sequences.

Researchers chemically synthesised 285 AI designed viral genomes and found that 16 successfully assembled into functioning viruses capable of infecting E. coli bacteria and reproducing.

The work focused on bacteriophages, viruses that infect bacteria rather than humans. Scientists involved in the research said the viruses were not harmful to people and that the model was not trained on viruses known to infect humans.

The study builds on decades of virus synthesis research, where scientists have used the genetic sequences of existing viruses to study disease mechanisms and develop treatments. What has drawn attention in this case is the AI model’s ability to generate large numbers of new genome sequences that appeared biologically viable.

The researchers trained Evo on more than 9 trillion nucleotides drawn from a vast collection of genetic sequences across plants, animals, microbes and viruses. They then focused on the Phi X-174 bacteriophage, a virus that has been studied for more than a century. From hundreds of thousands of proposed genomes, the team selected a small number for laboratory testing.

One practical finding from the study was that some AI designed phages were able to overcome bacterial resistance, a result that could have implications for phage therapy, an area of research aimed at treating antibiotic resistant infections.

The findings have also prompted debate about oversight and biosecurity. Some experts argue that AI could accelerate beneficial research in medicine and biotechnology, while others say the technology may outpace existing safeguards and regulatory frameworks.

Dom’s comments reflected the wider concern that AI is beginning to influence fields beyond software and automation, extending into biology and the design of living systems.

The study’s authors have emphasised that creating simple bacteriophages is very different from designing viruses that affect humans, and they have framed the work as a controlled scientific investigation. Even so, the research has become part of a broader discussion about how quickly AI capabilities are advancing and how societies should respond.

For supporters of AI driven scientific discovery, the work shows how machine learning can uncover patterns in biological data that would be difficult for humans to identify manually. For critics, it raises questions about governance, transparency and the limits of powerful generative technologies.

Dom’s observation that “intelligence is a tool” captured a point increasingly echoed across the technology sector: AI is no longer confined to analysing information, and is beginning to take a role in designing systems in the physical world, including biological ones.