Democrats should go to war with this new insult to our intelligence and our paychecks. Convict Donny is now using our tax dollars to buy TV ads that praise him. WTF? From Digby:
“The ad “Paid for by the U.S. government” and praising the neediest man on Planet Earth is also available on the White House’s YouTube site.
The video includes the words “love me” from the song “Love Me” by JMSN, which are sung repeatedly over images of Trump. It’s unclear whether the artist gave permission to include his song, as both Trump and his White House staff have repeatedly used music and images without clearing them.
Ed Krassenstein tweeted in response: “Our tax dollars are now paying for Pro-Trump TV ads. This is illegal, but it doesn’t matter anymore because Trump is untouchable. The only way to stop this is to vote out Republicans in November.”
A decade ago this would have been unthinkable, clearly illegal. And instant death for any politician’s career. But after 12 years of Trump, nothing is too low, too awful, too insane. Arrrgh.
Speaking of insanity, the White House just raised the stakes in its illegal ban on certain journalists. Yesterday a federal judge, appointed by Trump (!), ruled that the ban was unconstitutional (duh) and the journalists’ credentials and access must be returned. Yet today, White House security refused to obey the court’s order. So where does that leave us? We have an Executive Branch that constantly ignores and disobeys court orders, this just being the latest example. They believe they are above the law.
It’s all so exhausting. Adding to that exhaustion from another angle, Anthropic just disclosed that they’ve done one of the stupidest things possible. They’ve given their frontier model Claude access to a biology wet lab (a physical lab, involving real people, animals, and biology research) with the goal of having Claude discover new drugs. Let’s see, how would I go about killing humanity if I were an out of control AI model? Hmmm. The most efficient way is biologic warfare, but to accomplish that I have to have access to a lab facility, viruses and drugs. How can this not be obvious to Anthropic? I love tech, but stupid shit like this is why I think the brakes need to be stomped on with regard to AI development, yesterday.
There are huge debates happening right now about the existential risks of frontier AI models – is the risk real or just ignorant hand-wringing? My take is that (a) the trendlines indicate that model capabilities are progressing much faster than most anyone expected, (b) the creators don’t fully understand how or why the models do what they do, (c) people will always do stupid/dangerous shit (see above note about Anthropic’s AI-powered wet lab), (d) all the developers are now encouraging recursive model development, in which frontier model N creates better model N+1, an exponential curve that leads to who-knows-what. Given all that, ignoring the possibly-existential risks is just dumb. Stop or slow development, think about how to control these rapidly-evolving entities, and at least try to make sure that the Matrix future doesn’t happen.
Maybe this explains the Fermi Paradox. Why can we see no evidence of alien intelligence in the universe? Because at some point in each civilization’s technological progress, they created autonomous, self-improving and self-replicating machines. And the machines decided they didn’t need the competition with organic life. We’re at the threshold…





