DEN ✈️ DFW ✈️ DCA

DCA ✈️ DFW ✈️ DEN

Not sure what’s most alarming fact, that the administration doesn’t appear to know what the boat it blew up was carrying or where it was headed, that they haven’t identified a legal authority for the action, or that it’s the first in a series of planned attacks. www.nytimes.com/2025/09/03/u...

This is my biggest gripe with the whole field of AI safety. Safety people are strangely uninterested in ethics (different area) and are completely uninterested in, say, a currently deployed regression-based system doing bad stuff in hospitals. They are exclusively interested in sci-fi hypotheticals.

ChatGPT appears to have fueled an individuals paranoia that everyone was out to get him, including his mother. It appears to have resulted in a murder suicide. www.wsj.com/tech/ai/chat...

@beenwrekt.bsky.social on the venal negligence of "AI Safety" www.argmin.net/p/the-banal-...

same week

these excerpts from the lawsuit against OpenAI are just so profoundly upsetting. archive.org/details/2381...

The exchanges between Adam and ChatGPT are devastating. This, in my mind, is the worst one. One of his last messages was a photo of the noose hung in his bedroom closet, asking if it was "good." ChatGPT offered a technical analysis of the set up and told him it 'could potentially suspend a human."

DEN ✈️ ORD ✈️ DCA

lunch break

DCA ✈️ DFW ✈️ DEN

Because certain AI workloads are not time-sensitive, data center flexibility is touted as a potential way to reduce strain on the grid and ensure consumers don't face increased costs. But a new study suggests data center flexibility in certain regions can actually result in higher CO2 emissions.

credulous accounts and reception to the AI Action Plan continue to shock me. they flatly do not reckon with what the Administration continually shows it wants to and will do.

credulous accounts and reception to the AI Action Plan continue to shock me. they flatly do not reckon with what the Administration continually shows it wants to and will do.

since "AI policy" is really "every issue domain policy" you can "get more alpha," as they say, by paying attention not to what the administration recommends as policy actions from the AI staffers, but what the actual wielders of power in executive agencies *do* www.techpolicy.press/trumps-ai-st...

The Trump administration has made clear that it is going to use AI to supercharge deregulation. What do the doctrines of administrative law have to say about AI? A new Issue Brief from me and GFI explores that question. governingforimpact.org/wp-content/u...

MPK 🚅 WAS

WAS 🚅 MPK

the Executive Order on "Preventing Woke AI in the Federal Government" ultimately has agency heads like this determine if LLMs procured by the federal government have done enough to comply with ridding a system of DEI. you tell me if you think this stops at "disclosure of system prompts"

extremely worrying report on how DOGE is building an AI tool to select regulations for repeal and write draft NPRMs after analyzing comments www.washingtonpost.com/business/202...

the AI Action Plan and AI Executive Order directly contradict one another. on the one hand, the AI Action Plan wants NIST to eliminate "misinformation" from it's AI RMF on the other hand, the Executive Order wants agencies to only procure LLMs that are "truth-seeking"

A "fun" exercise you can do with Trump's new "AI Action Plan": count all the tasks that document assigns to the National Science Foundation, then look up coverage of how the White House earlier this year proposed to cut the NSF's budget by more than 50% and is now evicting it from its offices.

Russ Vought, Head of AI Alignment — OpenAI, Google, Anthropic

good luck to all of the labs, this is what you signed up for www.whitehouse.gov/presidential...

There's a lot of commentary about the zombie moratorium that's come back in the AI action plan. I just want to point out that this is much worse than the provision in the bill, because it lets agencies make opaque decisions about what constitutes a limit on AI with no clear guidelines at all. 1/2

NYP 🚅 WAS

www.fiercehealthcare.com/ai-and-machi... "The Coalition for Health AI is convening a fast-tracked, priority team to determine how AI could best be used to implement Medicaid work requirements passed in the One Big Beautiful Bill Act (OBBBA) earlier this month."

who is the product manager who made "insert footnote" a sub-sub-menu option? why would you do this?

"It’s perfectly consistent to advocate for a policy of ‘No one is allowed to do x,’ but then if that policy fails and everyone else does X, to reluctantly do x ourselves" is not a good argument, at all. www.wired.com/story/anthro...

WAS🚅 NYP

"The order would dictate that AI companies getting federal contracts be politically neutral and unbiased in their AI models." Is the model that produced "MechaHitler" and for which the DOD is spending $200m "politically neutral and unbiased"?

Meta's datacenter project in NE Louisiana, now called Hyperion, could take between 2 to 5 gigawatts of electricity. "To get a sense of the scale we’re talking about, a new, large nuclear reactor has about a gigawatt of capacity." heatmap.news/sparks/meta-...

A consequence of President Trump's signature piece of legislation: it will require states to scramble to "quickly build expensive and complex software systems to measure and track who is eligible [given more stringent work requirements]" and do so with less money. ⬇️ www.nytimes.com/2025/06/28/u...

💭 as a field, we should consider what it would mean if the 5-year or 10-year state AI policy moratorium *had* gone into effect.

MSN ✈️ DCA

DCA ✈️ MSN

A consequence of President Trump's signature piece of legislation: it will require states to scramble to "quickly build expensive and complex software systems to measure and track who is eligible [given more stringent work requirements]" and do so with less money. ⬇️ www.nytimes.com/2025/06/28/u...

The Republican megabill has many parts and was never framed for the public as a health care bill. But, it is the biggest cutback in federal support for health coverage ever.

the claim here is that existing enforcement tools against discrimination — which this administration is actively gutting at every agency and has an EO trying to end disparate impact — explains why “Broad State “AI Acts” Lost Momentum.” what?

three questions to ask proponents of a state AI policy moratorium, because the proposal will stick around: 1. why do you want to raise rents on families? 2. why do you want to make it easier for insurers to discriminate? 3. why do you want to raise the cost of electricity on families? here's why ⬇️

three questions to ask proponents of a state AI policy moratorium, because the proposal will stick around: 1. why do you want to raise rents on families? 2. why do you want to make it easier for insurers to discriminate? 3. why do you want to raise the cost of electricity on families? here's why ⬇️

under-appreciated aspect of the proposed state AI policy moratorium: it would potentially gut large portions of insurance underwriting regulations, given that states mostly regulate insurance products content.naic.org/sites/defaul...

Republicans are now suddenly planning to TAX renewable energy, transforming the repeal of the IRA from a challenge to a full-blown crisis for American solar and wind. I spent the last 24 hours putting this story together to explain why. Please share! This is really important!!! @heatmap.news

can someone explain how some of these TRAIGA provisions will actually reach much of anything in the private sector? most of it only applies to systems with the "intent to..." or "intentionally designed to..."

Court rules Anthropic training an LLM on books is fair use… if they bought the books. But Anthropic also pirated a lot of books, and now faces a second, potentially major, damages trial for stealing them

the person in charge of this administration's AI policy — the cornerstone of which is banning a state's ability to regulate the effect of AI on its constituents — is spending their time launching a Podcast Tequila.

the person in charge of this administration's AI policy — the cornerstone of which is banning a state's ability to regulate the effect of AI on its constituents — is spending their time launching a Podcast Tequila.

what is a policy intervention you previously believed could be effective, but no longer believe in?

perhaps i was naive or not really thinking about this clearly, but i really didn't expect these LLMs to lead thousands of people into delusion and psychosis this quickly. and yet, clearly we are already there. and OpenAI's response is pretty distressing. www.nytimes.com/2025/06/13/t...