“American presidential candidate” - I’ll define this as the set of candidates for the 2028 presidential election who participate in at least one primary debate for either mainstream party, as well as candidates (from any party) who at some point after January 1, 2028 are polling at above 3% in at least two reputable general election polls.

“A platform that includes nationalizing or pausing AGI” - Issues related to AI do not need to be the primary focal point of the candidate platform, but the candidate must regularly reference in their public appearances a stance on AI that directly and explicitly supports a significant form of nationalization of either development or deployment. Here are some roughly outlined stances which would/would not count:

Would count towards a YES resolution:

• “AGI development cannot continue to be done by private companies who are not directly accountable to the American people”

• “There must be a Manhattan Project for AI”

• “We must pause AGI development [or large model training] until we can ensure the technology is safe” (possibly except for government projects)

• “The American government must assert complete ownership of any [sufficiently large/capable] model trained by an American company”

• “[Large AGI project/company] must come under Federal government control”

• “Once [some clear threshold in capabilities] is reached, [any of the above]”

Would not alone count toward a YES resolution:

• Policies on narrow AI platforms

• Support for a requirement that sufficiently capable models must be approved/vetted by the government before commercial deployment, without asserting ownership of the IP

• “The US government should have its own national AGI project” (without integrating existing AGI companies)

• Vague statements like “We must prevent AI from causing mass unemployment” without mentioning some form of nationalization

I’m happy to discuss further delineations of these criteria.