'So striking': Trump's 'little robot' quip amid AI uproar alarms analyst
President Donald Trump is interviewed by Reuters White House correspondent Steve Holland (not pictured) during an exclusive interview in the Oval Office in the White House in Washington, D.C. on Jan. 14, 2026. REUTERS/Evelyn Hockstein

President Donald Trump is all-in on expanding artificial intelligence development — and completely dismissive of the risks, in a way that left liberal analyst David Pakman alarmed.

This comes after a number of disturbing incidents, including a cyberattack masterminded directly by a rogue ChatGPT model that escaped the lab — as well as a growing chorus of AI experts warning that if left unchecked, the technology could wipe out the human race.

"Guys, I'm starting to think that Donald Trump might not know anything about AI," said Pakman. "And this is not like a ha-ha story. This is scary, because he is confidently making declarations about AI and he doesn't know a damn thing about it."

Trump was actually confronted about the threat of AI going rogue, noted Pakman, and had a very simple response.

"It's going to be fine," said Trump in the clip. "We'll always have something to stop them, right? We'll have a little gear. I really hope — I don't like that. I really don't like that robot. We'll stop — a little gear, boom. We don't like the robot."

Trump's attitude, said Pakman, "would be funny if we were talking about robot vacuums — if this was about your Roomba, and you don't like how it's cleaning your carpeted bathroom, and you're like, 'You just do a little gear, we don't like this robot.' But this is the president of the United States talking about how humanity would maintain control over potentially superhuman artificial intelligence. And his answer is, basically: "There's a little gear; I don't like that robot."

And it's so striking because some of the people who actually built these systems are asking governments, "Hey, we need stronger safeguards." It could be self-serving — I'm open to that — but you've got OpenAI and Anthropic and others saying we need independent evaluations, we need a national AI safety council, all of this stuff. Google DeepMind's leadership called for a global AI watchdog. And Trump's response isn't "little gear, boom, I don't like that robot."

This is all happening as Trump plays around with ChatGPT and uses it to explain laws and policies that he is already doing and should understand as president, Pakman noted.

"Trump doesn't understand far simpler things than AI," he said. "This is something technically complicated and potentially consequential, and Trump has the complete and total confidence he has always had ... That's dangerous. And if for no other reason — even though there are many other reasons — if for no other reason than to rein in a little bit of this belligerence, we've got to take power away from Republicans in November, at least in the House, if not the House and Senate."

- YouTube www.youtube.com