Earlier this month, Shayna Korol tracked a move by AI leaders and biosecurity experts to secure DNA synthesis machines, treating them as physical chokepoints against catastrophe. It’s a practical, high-stakes attempt to build hard boundaries around what a machine can do. But writing just a few days later, Julio M. Ottino looks toward Leo Tolstoy to suggest that the real danger might be the feeling of safety itself. There’s a friction between these two perspectives. One seeks security through technical constraints, while the other warns that believing the future is predictable can dull the very human alertness needed to face it. It's worth pausing on whether these guardrails provide control or just the comfort of an illusion.

Korol presents a world where safety is a matter of finding the right chokepoint—specifically, the screening of genetic sequences by commercial providers. There is a certain comfort in the idea that if the CEOs of OpenAI and Google DeepMind can agree on a set of technical guardrails, the path toward AI-assisted catastrophe might be blocked. But read alongside Julio M. Ottino, that focus on preventative architecture starts to look like the exact dream of certainty that Tolstoy warned against. While Korol points toward the need for a hard science of prediction to stop bad actors, Ottino suggests that the most vital human quality is not the perfection of a plan, but what he calls alertness. The interesting thing is that these two perspectives don't just disagree on methods; they disagree on the nature of reality itself. Korol assumes the future can be managed through the right code and the right regulations. Ottino, however, points toward Prince Andrei on the eve of battle, where outcomes hinge on fleeting, unrepeatable moments. Putting the two together suggests that the real risk might not just be the technology itself, but the belief that human agency can be replaced by a system designed to engineer away the inherent randomness of what comes next.

Screening DNA sequences for pathogens provides a concrete lever to pull, a way to keep the lab bench from becoming a site of danger. It offers the comfort of a technical fix. Still, Tolstoy’s battlefield suggests that history remains a series of accidents. Relying on predictive systems to manage risk might actually dull the sharp alertness required to face a chaotic world. If the focus remains on perfecting the code, how does one stay ready for the things that no algorithm can see coming?

Sources:
Vox: The next AI safety fight may actually be about DNA
The Hedgehog Review: Tolstoy and the Illusion of Inevitability