# Defensive Co-Scaling **Entity class:** Governance strategy **Domain:** Artificial intelligence / science / political economy **Maturity:** Developed ## Definition **Defensive Co-Scaling** is the principle that protective capacity should grow with the system it protects instead of being used as a reason to prevent the system from growing. ## Mechanism and Consequence The city is the governing analogy: population created new risks, but police, fire protection, sanitation, and public health scaled with urban life. Advanced AI requires the same coupling of capability with monitoring, security, alignment, and institutional response. ## Relationships - [[wiki/AI Control|AI Control]] - [[wiki/AI Acceleration|AI Acceleration]] - [[wiki/Public Utility AI|Public Utility AI]] - [[wiki/Risk Debate|Risk Debate]] ## Source Dossiers - [[research/AI Is Now So Good They Will Pay You to Quit|AI Is Now So Good They Will Pay You to Quit]] - [[research/Super Intelligence Week Index of Statements from the White House Luncheon|Super Intelligence Week Index of Statements from the White House Luncheon]] - [[research/Jensen Pushes Back on Doomers Moonshots Live|Jensen Pushes Back on Doomers Moonshots Live]] ## Simple Reminders, Quotations, and Thoughts > "Defensive co-scaling is how we end up in a great future. At the dawn of cities, it would have been a terrible idea to say we could not form cities because cities breed crime. Cities are fountains of economic growth. Instead, police forces, fire departments, and other municipal services defensively co-scale with the population. The singularity will need exactly the same thing: do not decelerate; defensively co-scale." > **— Alexander Wissner-Gross**, *MOONSHOTS Live, September 25, 2026* [[reminders/AI Control/The Singularity Needs Defensive Co-Scaling, Not Deceleration by Alexander Wissner-Gross|The Singularity Needs Defensive Co-Scaling, Not Deceleration by Alexander Wissner-Gross]] > "If we all do a good job and act responsibly, [AI is not going to kill us]. I am optimistic that this is going to be a very positive future for everyone. There is a lot of rhetoric filled with doom, but there are a number of labs working on advanced AI, and I do not think we need industry-wide coordination in order not to mess this up. Each lab needs to take the time, when it sees issues, to make sure it is proceeding safely. We are entering a phase where trust and alignment will be the most important next set of capabilities." > **— Mark Zuckerberg**, *MOONSHOTS Live, September 25, 2026* [[reminders/AI Control/Responsible AI Labs Can Build a Positive Future for Everyone by Mark Zuckerberg|Responsible AI Labs Can Build a Positive Future for Everyone by Mark Zuckerberg]] > "I think the worst possible outcome [for superintelligence] is deceleration, so don’t decelerate." > **— Alexander Wissner-Gross**, *MOONSHOTS Live, September 25, 2026* [[reminders/AI-Science Doomers/The Worst Outcome for Superintelligence Is Deceleration by Alexander Wissner-Gross|The Worst Outcome for Superintelligence Is Deceleration by Alexander Wissner-Gross]] > "We have evolved for four billion years fighting all sorts of threats. The universe tries to wipe you out; entropy tries to wipe you out. We are local anti-entropic phenomena that somehow manage to create order out of chaos, and the idea that we will now be wiped out by [AI] seems unbelievably absurd." > **— Salim Ismail**, *MOONSHOTS Live, September 25, 2026* [[reminders/Civilization/Humanity Is a Four-Billion-Year Anti-Entropic Phenomenon by Salim Ismail|Humanity Is a Four-Billion-Year Anti-Entropic Phenomenon by Salim Ismail]] > "A few decades ago, the United States government tried to ban encryption, and it was shown that banning encryption implicates free speech because you are trying to ban mathematics. Superintelligence is mathematics, and banning it should be unconstitutional because mathematics is free speech." > **— Emad Mostaque**, *MOONSHOTS Live, September 25, 2026* [[reminders/AI-Science Doomers/Banning Superintelligence Means Trying to Ban Mathematics by Emad Mostaque|Banning Superintelligence Means Trying to Ban Mathematics by Emad Mostaque]] > AI-doom scenarios grant attackers near-magical abilities while leaving defenders frozen in the present. If intelligence can design a perfectly concealed supervirus, the same class of capability can design extraordinary detection and vaccines. Forecasts that scale only the attack are not balanced forecasts. > **— Adapted from Richard Socher**, *MOONSHOTS Live, October 2026* [[reminders/Risk Debate/Doom Scenarios Give Attackers Superpowers and Defenders Nothing by Richard Socher|Doom Scenarios Give Attackers Superpowers and Defenders Nothing by Richard Socher]]