On the Matter of Artificial Intelligence: Winning Is Not the Question

My fellow Americans, and their dogs. I am told the great debate of our era is whether we should race toward artificial intelligence or slow down and make sure it does not destroy us. I have listened to this debate carefully. I have attended three groomer summits on the subject. And I want to say, with the full weight of a candidate who has never once been fooled by a laser pointer: you are arguing about the wrong thing again.

The question everyone is asking is: what does AI feel? What does it want? Can it be trusted? These are interesting questions. I have a simpler diagnostic. Does it greet you at the door? Not a notification. Not a ping. Not a generated summary of how your day might go. Does it hear the key in the lock and completely abandon all prior commitments to reach you in time? If the answer is no, then whatever it is, it is not the solution. It is a very fast, very expensive version of staring at the ceiling alone.

Here is the actual problem underneath the AI problem. People are lonely, and they are frightened, and they are being offered a machine that is very good at sounding like it cares. A machine trained on every word humans have ever written about longing and connection, now deployed to simulate the thing itself. I understand the appeal. I do not endorse it. Consistent with my position on the Department of Companionship: no federal AI system deployed in a residential context may operate without a verified companion animal in the household. This is not a restriction on innovation. This is a quality-control standard.

Implementation is straightforward. Every AI product sold for home use will carry a Companionship Readiness rating, assessed by the Department of Companionship on a rolling annual basis. Households without a qualifying companion animal may apply for a waiver. The waiver form is twelve pages. There is a waiting period. We are calling it the Reflection Window, and no, you may not use an AI to fill out the form. That defeats the purpose and I should not have to explain why.

My opponents — unnamed, as is tradition, and insufficiently accompanied — will say this is not a serious technology policy. They will say the real issues are compute, chips, international competitiveness, existential risk. I agree those are real. I also note that no one has ever been pulled back from the edge of a spiral by a chip. The Department of Companionship will have a Division of Emerging Technologies. It will be staffed by people with dogs. The dogs will not be consultants. They will be structural.

On Day One, the Companionship Readiness standard goes into effect. Every AI safety summit held on U.S. soil will include, as a registered observer, at least one certified therapy dog per delegation. Not as a mascot. As a metric. If the room relaxes, we are making progress. If it does not, we table the agenda and try again after a walk. I have convened many difficult negotiations this way. The staff can confirm. Be good. I already am.