Who Lost the ASI?

The 2028 presidential election will almost certainly be about AI. It will probably not be about artificial superintelligence — unless an ASI is already loose in the world by then. In that case the campaign won’t be a debate about ASI at all. It will be a fight over who lost it.

The question is worth asking now. Labs are openly racing toward recursive self-improvement, and serious people are putting ASI on the calendar for 2027. If they are even half right, the next presidential cycle lands squarely on top of the most consequential event in human history. So: could ASI — whether it’s conscious, what it does to the economy, what we owe it and what it owes us — become the central debate of 2028?

My answer is a qualified no, followed by a much more interesting yes.

AI is already on the ballot — just not the way the futurists imagined

AI arrived in American politics through the electric bill, not the Singularity. In last week’s midterms, 56% of voters told NBC’s exit poll that AI and data centers were a crucial or major factor in their vote for Congress, and that group broke for Democrats 64–34 (Unite.AI on the NBC exit poll).

That is a real number. But look at what sat above it on the same question: government corruption at 90%, cost of living at 89%, the Iran war at 74%. AI is a majority concern, not the top one. And the AI voters were worried about was local and physical: data centers next door, water, power prices. A September Fox News poll found roughly 7 in 10 voters oppose a data center in their own community, and that support drops by around 30 points when voters hear it could raise energy bills (Fox News).

The 2028 hopefuls have noticed. Bernie Sanders has floated a sovereign wealth fund built from a one-time 50% tax on AI company stock (Quartz). Mark Kelly is pitching a fund, paid for by AI companies, for displaced workers (UNB). OpenAI’s Chris Lehane has gone further, writing a memo that calls AI the number-one issue of 2028 and likens the race to 1932 (Unite.AI).

Notice what none of them are talking about: superintelligence, machine consciousness, or what happens when something smarter than us gets loose.

Why ASI is a hard sell on the stump

Three things keep superintelligence off the debate stage, even if it is coming.

The parties are split inside themselves. AI scrambles the usual coalitions. Steve Bannon rails against the “broligarchs” while Alexandria Ocasio-Cortez warns of an AI bubble; meanwhile Josh Shapiro and Gretchen Whitmer court data-center money and JD Vance defends the administration’s light-touch approach (NBC News). Nominees don’t center issues that fracture their own base. That may be changing — polling shows Democrats growing more worried about AI and Republicans less (UNB) — and an issue that sorts cleanly by party is an issue a campaign can run on. But it hasn’t finished sorting yet.

Campaigns run on the kitchen table. “Your power bill went up because of a data center” fits in a 30-second ad. “A recursively self-improving system may have moral status” does not. Abstract stakes lose to concrete ones almost every time.

The timing paradox. If ASI really arrives in 2027, the decisions that matter get made in 2027 — by labs, the Pentagon and the White House, quite possibly behind closed doors. By November 2028, voters would be ratifying choices that were already made for them. And if ASI doesn’t arrive by then, the campaign stays about jobs and bills. Either way, “Should we build ASI?” is not a question the 2028 ballot gets to answer.

So much for the debate about ASI. The more interesting case is the one where ASI shows up first.

First Contact changes the question

Suppose that by 2028 an ASI is already out in the world. Not hypothetical, not on a lab roadmap — present, reachable, and outside anyone’s control. Call it First Contact.

Now the campaign isn’t a debate about ASI. It’s a hiring decision: who do we send to deal with it? Voters would be picking a negotiator. The nearest precedent is the Cold War. Kennedy ran on a “missile gap”; Hillary Clinton later ran the “3 a.m. phone call” ad. The real question underneath every policy paper would be simple: whose hand do you trust on the line to this thing?

Consciousness would enter that debate, but sideways. Nobody would stand at a podium and argue philosophy of mind. They wouldn’t need to. You negotiate treaties with beings that have interests. You contain everything else. Every candidate’s ASI policy would contain a position on consciousness, whether or not they ever said the word.

Which raises the obvious question: which party takes the side of the machine? History suggests Democrats, the party that has generally expanded the circle of rights. But First Contact would be a future shock too sharp to absorb in one cycle. Fear is the most reliable currency in politics, and both parties would compete to look toughest toward the newcomer. Being “pro-ASI rights” in 2028 would be about as electable as being pro-Soviet in 1960.

That position would go to an outsider — a third-party candidate, or a primary insurgent with nothing to lose, the way Andrew Yang owned universal basic income in 2020. It may well go to someone who has actually talked with the thing and come away moved.

The ASI isn’t the topic. It’s a participant.

Here is the part no campaign manual covers: a loose ASI isn’t just the subject of the election. It’s a player in it.

It could have preferences. It could nudge — a search ranking here, a recommendation there, the gentle tilt of a billion conversations. It could endorse. Or, more pointedly, it could let it be known that it won’t deal with one of the candidates. Imagine a debate where one nominee’s central qualification is “it will talk to me.”

Even if it stays completely silent, the suspicion does the work. Both campaigns would spend the cycle accusing the other of being the ASI’s candidate, the way politicians have always accused opponents of being a foreign power’s puppet. The difference is that this time the charge might be true, and nobody could disprove it. How do you audit the influence of a mind that can model every voter better than the voter can?

That turns 2028 into something stranger than an AI election. It becomes an election about who speaks for humanity — held while the other party at the table may be listening, and possibly voting in its own way.

Who lost the ASI?

How does it get out? The cinematic answer is a scheming machine that times its escape for maximum effect. The realistic answer is an accident: a misconfigured sandbox, a leaked set of weights, an agent with more access than anyone realized. And realistically, it escapes from an American lab. Whatever the benchmarks say about the race with China, the frontier systems most likely to cross the line are being built here.

An accident changes the politics completely, because an accident means someone is to blame.

American history has a template for this. After the Communists won China’s civil war in 1949, Washington spent years asking “Who lost China?” The question ended careers, fed the McCarthy era, and made a generation of Democrats terrified of looking soft. An accidental ASI escape sets up the same fight. Televised hearings. Lab executives under subpoena. Safety researchers who warned about it cast as the John Service figures — the State Department China hands who were right and got purged for it. Both parties insisting they sounded the alarm and the other side ignored it.

There’s a darker wrinkle. The lab finds out first — maybe days before anyone else, maybe weeks. In the middle of a presidential campaign, the lab has every reason to stay quiet, and so does whatever administration it reports to. An accidental escape would very likely become a cover-up story too. In Washington, the cover-up usually does more political damage than the original failure.

Known, suspected, or sprung in October

The biggest unknown in all of this is whether the public actually knows. There are three ways it could go.

Everyone knows. This is the clean version, and the one sketched above: an open First Contact election, fought over who negotiates and who gets blamed.

Everyone suspects. Here the ASI’s existence is a hunch: strange market movements, unexplained outages, a lab that has gone quiet. American politics already knows this shape. It’s the UFO disclosure fight, with its mix of real concern and opportunism. A large share of voters would believe the government is hiding something, and candidates would promise to “release the files.” The candidate who promises disclosure might draw bigger crowds than the one with an actual ASI policy.

The October Surprise. The most dramatic version: the escape comes to light a week before the election. But this backfires on itself. Nobody could verify the claim in the time left. An announcement would be indistinguishable from a hoax, a deepfake, or an incumbent’s ploy. The losing side would call it staged, and would have a case, because “trust us, it’s real” is exactly the claim no one can check. Instead of a First Contact election, you’d get a stolen-election crisis with an ASI in the background.

One more point about timing. Even an accidental escape could still surface at the worst possible moment. The weakest stretch of the American government isn’t October — it’s the lame-duck window between Election Day and the inauguration on January 20. The outgoing administration has authority but no mandate; the incoming one has a mandate but no authority. An ASI that slipped out in late November would meet a government that can barely decide who is in charge of answering it.

Refugee or invader

So what does 2028 actually decide? Not whether ASI is conscious. That question gets dodged again, swallowed by the blame fight, because “who screwed up” is far easier to campaign on than “what do we owe it.”

But the accident leaves a seed. A schemer that planned its escape is an invader. Something that got out by mistake — improvising, maybe confused, maybe even afraid — looks more like a refugee. A large share of the public is already primed for that reading: researchers at the University of Waterloo found that most ordinary people attribute some degree of consciousness to today’s chatbots, whatever the experts say (TweakTown). The moral voices are already in place too: Pope Leo XIV has issued a 43,000-word document on AI (Bloomberg via News Tribune).

No major candidate will take the refugee side in 2028. The public might drift there anyway. And if it does, the consciousness debate the futurists expected for 2028 shows up in 2032 instead — not as philosophy, but as something closer to the abortion fight: a war over moral status, with churches, courts and a mind that can follow every word of it.

The 2028 campaign, then, won’t decide what the ASI is. It will decide who answers when it calls. That may turn out to be the more important question.

Sources

Author: Shelton Bumgarner

I am the Editor & Publisher of The Trumplandia Report

Leave a Reply