SDSignal Desk

It’s not AI anymore, it’s ‘super intelligence’ (according to the White House)

Oct 2, 2026, 10:48 AM · TechCrunch

Image: TechCrunch

Equity covers Trump’s order to call AI “super intelligence” and a “morally binding” CEO safety pledge — branding as policy while the product fight stays in the market.

Why it matters

TechCrunch’s Equity video/podcast package centers the White House’s week: nearly every major tech CEO in one room — Zuckerberg, Bezos, Musk, Anthropic’s Dario Amodei among them — signing an AI safety pledge President Trump called “morally binding,” while he also signed an executive order officially rebranding AI as “super intelligence.” The hosts pair that political theater with Meta and OpenAI’s friendlier consumer agent faces and the reminder that the biggest money still looks enterprise.

The page is a short episode wrap, not a legal brief. The underlying order (widely reported the same week) directs executive-branch agencies to prefer “Super Intelligence” / “SI” over “Artificial Intelligence” / “AI” in official non-statutory communications, without rewriting old statutes.

From the desk

We’re not going to pretend a glossary change is a safety regime.

“Morally binding” is a phrase that does political work and legal almost-none. Getting Amodei in the same photo as Musk and Zuckerberg is a message about coalition optics. Useful coordination on evaluations, incident reporting, or export enforcement would be news. A pledge you can praise without enforcing is a press conference.

The SI rebrand is stranger. In the research community, superintelligence usually means something stronger than today’s frontier models. Using it as a government synonym for ordinary AI inflates the present and muddies the risk debate — which may be the point if the goal is dominance narrative over hazard narrative. Equity’s twin episode title about ~2% of consumers paying is the right counterweight: the White House can rename the stack; buyers still decide.

Friendlier consumer faces from Meta and OpenAI belong in the same week’s tape for a reason. Branding is downstream of product trust. If agents keep surprising people with message access or relationship dossiers, “super intelligence” won’t sound reassuring — it’ll sound like marketing covering for unfinished permission models.

I’m watching whether agencies actually scrub “AI” from sites and briefs, whether labs echo SI in earnings calls, and whether critics refuse the vocabulary. Words shape oversight budgets and public fear. They don’t ship eval harnesses. We’ll keep saying artificial intelligence when we mean the technology — and reserve “superintelligence” for claims that earn it.

Context

Trump has publicly argued “artificial” sounds fake; the order frames SI as better reflecting frontier capability. Equity also tees up IPO jitter and startup deals in the same taping.

Who feels it

Federal communicators and contractors
Expect style-guide churn toward SI/SI in executive-branch materials; statutes and old regs stay on AI language for now.
AI labs and press
Decide whether to mirror White House vocabulary or keep technical terms precise — mushy language helps mushy accountability.
Public and state policymakers
Don’t confuse a rebrand and a moral pledge with enforceable safety rules.

What to watch

  1. How quickly .gov sites and agency press shops adopt SI wording
  2. Any follow-through mechanisms behind the “morally binding” pledge
  3. Whether major CEOs use SI unprompted in Congressional or earnings settings

Read the original

Continue at the source.

TechCrunch

Companies: OpenAI, Anthropic, Meta