The more things change, the more they stay the same.
asserted
they → change → ?
Yesterday, Trump dusted off his old attacks against former Fed Chair Jerome Powell: “He should be forced to resign, IMMEDIATELY!
asserted
He → dust → Powell
If he doesn’t resign, he should be sued, at the highest level, by the United States Government, for either corruption or incompetence.”
asserted
he → resign → corruption
It’s all great news for new Fed Chair Kevin Warsh: He hasn’t been able to lower interest rates yet, either, but for now, Trump seems content to keep blaming Powell.
asserted
Trump → ’ → Powell
It’s not just that a new Quinnipiac poll shows only 34 percent of voters approving of Donald Trump’s job performance, with 61 percent disapproving.
asserted
percent → ’ → performance
It’s that Quinnipiac has Democrats leading by 12 points on the generic congressional ballot.
asserted
Democrats → ’ → ballot
The FiftyPlusOne polling aggregator shows “only” about an 8-point generic ballot lead for the Democrats.
asserted
aggregator → show → Democrats
But even that’s a ten point swing from 2024, when Republicans won the national vote for the House by 2.6 points.
asserted
Republicans → ’ → points
A 10-point swing would be a wave that puts lots of districts and states in play.
asserted
that → put → play
A 14-point swing, which is what Quinnipiac implies, would be a tsunami.
asserted
Quinnipiac → imply → what
And I’ll note that Quinnipiac has been on the money before.
asserted
Quinnipiac → note → money
Their poll in late September 2024 had Trump leading Kamala Harris by 48 percent to 47 percent.
asserted
Trump → have → 47
I remember seeing those numbers and hoping against hope that they were wrong, that surely Harris had at least a narrow lead.
asserted
Harris → remember → lead
Just as notable as the generic ballot in Quinnipiac’s poll this week was what it found about voters’ views on how our government should deal with the challenge of artificial intelligence.
asserted
government → find → intelligence
The results were, to say the least, lopsided.
asserted
results → say → least
“Seventy-one percent of voters say they are more likely to vote for a political candidate who supports stricter guardrails for AI development,” compared to 16 percent who preferred a candidate who supports expanding AI development.
asserted
who → say → development
Even Republican voters preferred stricter guardrails, by 52 percent 29 percent.
asserted
voters → prefer → percent
What’s more, 79 percent of respondents said that with respect to AI, they think safety is more important than staying at the forefront of innovation.
asserted
safety → ’ → innovation
Only 15 percent thought the opposite.
asserted
percent → think → opposite
Again, even Republicans prioritized safety over innovation, by 62 percent to 32 percent.
asserted
Republicans → prioritize → percent
Voters seem to have noticed that the Trump administration doesn’t agree with them.
asserted
administration → seem → them
Only 25 percent support the president’s handling of the issue, while 58 percent disapprove.
asserted
percent → support → issue
And that disapproval has started to spill over to his party.
asserted
disapproval → start → party
Forty-two percent now prefer the Democratic party on AI, while 29 percent prefer the Republican party.
asserted
percent → prefer → party
Only three weeks ago, by contrast, a New York Times/Siena poll found basically no gap between the two parties on the issue.
asserted
poll → find → issue
I’ll add that this survey was in the field over the weekend, before Tuesday’s White House lunch for the tech oligarchs.
asserted
survey → add → oligarchs
There Trump confirmed he was going to do nothing to pause or safeguard or regulate developments in artificial intelligence.
asserted
he → confirm → intelligence
But Trump did issue an executive order rebranding “Artificial Intelligence” as “Super Intelligence.”
asserted
Trump → issue → Intelligence
The order celebrated the fact that these technologies are “transformative” and will allow us to do “what was previously impossible.”
asserted
what → celebrate → fact
But aren’t “transformative” technologies that allow the “previously impossible” the kind of thing that requires some oversight by government, and some guardrails that depend on more than the goodwill of tech oligarchs?
asserted
that → allow → oligarchs
Indeed, it turns out that the “White House Accord on Super Intelligence” released after the lunch of the AI oligarchs was drafted by . . . the AI oligarchs.
asserted
Accord → turn → oligarchs
As Semafor reports:
asserted
Semafor → report → ?
The idea for what became President Donald Trump’s Tuesday pact with the AI industry came from a conversation between Meta CEO Mark Zuckerberg and House Speaker Mike Johnson at last week’s state dinner . .
asserted
what → become → dinner
Zuckerberg spoke to Johnson—whom he was seated next to at the dinner for Chinese leader Xi Jinping—about regulatory concerns around AI . . .
asserted
he → speak → AI
The Meta CEO then talked to Nvidia CEO Jensen Huang about crafting a set of principles that tech executives could use at their next meeting with Trump . . .
Zuckerberg then circulated a draft ahead of Tuesday’s meeting between Trump and AI executives.
uncertain
Zuckerberg → talk → Trump
Huang played a role gathering support for it among the administration’s guests . . .
asserted
Huang → play → guests
Eventually Trump himself released the document as a “White House Accord on Super Intelligence,” outlining broad principles for self-regulation.
asserted
Trump → release → regulation
That story apparently comes from a single source, so it’s possible there’s more to it than that—but it certainly looks like the “White House” executive order was drafted by the AI firms.
asserted
order → come → firms
Trump merely signed it.
asserted
Trump → sign → it
It was nice of Trump to throw the glitzy lunch for the AI guys and sign off on their no-safeguards policy.
asserted
It → throw → policy
…and 41 more, not listed.