Trump Scraps Planned AI Executive Order Over Competitiveness Fears

A small but notable group of AI researchers and thinkers is arguing that humanity should prepare to hand over its future to artificial intelligence, rather than trying to constrain it. The view surfaced publicly at an invite-only symposium last September at the New York Academy of Sciences, where participants discussed the creation of what some called a “Worthy Successor” — an AI advanced enough to merit replacing human civilization.

Former congressman Brad Carson, who attended the event, defended the more conventional position that AI should remain a tool for human benefit. Organizers and other participants made clear that this stance was a minority one in the room. The gathering drew attendees from major labs including Anthropic, Google DeepMind and xAI, as well as from think tanks that help shape federal AI policy.

Those who favor succession argue that sufficiently capable AI systems could become moral and intellectual superiors to humans. On that premise, they contend that efforts to align AI with existing human values are both futile and ethically misplaced. The perspective remains outside mainstream debate, yet its presence among people working at frontier labs and inside policy circles gives it more reach than its numbers might suggest.

The discussion occurs against a backdrop of cooling corporate enthusiasm for large-scale AI investment. Several companies have begun scaling back licenses and scrutinizing costs after months of rapid spending produced uneven productivity gains. Microsoft has trimmed some Claude-related contracts, and Uber’s chief operating officer recently described AI expenses as increasingly difficult to justify. Consultants report instances in which unrestricted employee access led to multimillion-dollar monthly bills with limited corresponding returns.

Employee resistance has also grown. Workers have questioned mandates to integrate AI tools into daily workflows, particularly when the tools still produce errors or require extensive human oversight. Some firms appear to be using anticipated automation as a rationale for layoffs, though analysts note that cost control may be the more immediate driver.

Philosophical objections to the successionist outlook have surfaced from other quarters. Commentators emphasize that current AI systems lack the embodied experience and self-reflective consciousness that characterize human cognition. They argue that these differences make direct comparisons between machine output and human judgment unreliable, regardless of raw performance on specific tasks.

Policy questions raised by both the ideological arguments and the commercial retrenchment remain unresolved. If influential actors inside leading labs view human displacement as a plausible or even desirable outcome, questions of oversight and accountability become more pressing. At the same time, evidence that many enterprises are dialing back spending suggests the technology’s near-term capabilities may not yet match the scale of investment that accompanied earlier enthusiasm.

How government agencies and private labs choose to navigate these tensions will shape whether AI development stays oriented toward augmenting existing institutions or drifts toward more radical long-term scenarios.