And a lot of his predictions about the economics of these companies are factually true. They are burning money. It's just that they haven't run out of money to burn yet.
04.03.2026 19:25
👍 2
🔁 0
💬 0
📌 0
And a lot of his predictions about the economics of these companies are factually true. They are burning money. It's just that they haven't run out of money to burn yet.
I haven't extensively debugged this yet but it's fun. Play as Hubel & Wiesel and determine the RF properties of a V1 neuron.
cogneurolab.org/PSYC310/v1-r...
Yes! Yes!! Yes!!!
I thought the Expanse series had one of the most useful (tight, accurate, nonintrusive) versions of AI I've ever seen in a show.
But I'd love to see a version where the Rocinante had an multimodal LLM for the flight controls. Alex would just be screaming at it...
www.tiktok.com/@fatherphi/v...
Dr Eben Daggett and I have a book chapter coming out very soon. It's on modeling similarity using MDS, and it can be found in the upcoming 2nd edition of the Encyclopedia of Measurement in Social Sciences.
Claims of 40 dead after an Israeli airstrike on a elementary girl’s school in Iran
www.aljazeera.com/news/liveblo...
@GOP on X: Vote the pro-peace ticket. Vote Trump-Vance 🇺🇸 04 Nov 24
The lies about Trump's approach to war and foreign policy were ubiquitous. Just a few.
I think she read that 30% of the homeless population reside in California.
What she READ was that 30% of the homeless in the US reside in California. What came of her mouth was:
News media really thought they could dabble in a little propaganda to curry favor with wealthy warmongering control freaks instead of being tirelessly dedicated to truth & now it’s FAFO time.
It’s like how every now & then you find the bones of a Floridian family *inside* of their pet python.
A follow-up thread on Sam's essay about LLMs and epistemic vigilance.
TLDR: LLMs make epistemic vigilance impossible because they obscure sources.
A major Pentagon–Anthropic showdown is underway. The Pentagon is pushing for Claude to enable mass AI surveillance + autonomous drones, threatening a supply-chain-risk label or a defense production act takeover if they don’t.
Ultimatum was issued for tomorrow.
Anthropic just refused.
New York Times headline
This is a very misleading headline from the New York Times:
"Clinton Briefly Halts Epstein Deposition After Republican Leaks Photo."
In reality, her lawyers asked the committee to allow reporters in the room after Boebert leaked a photo of the proceedings.
This is deeply illegal, breaking the Medicaid statute and the ICA.
States are entitled to funding from federal govt to help pay for Medicaid. Providing less than the statute calls for is open and shut illegal, more akin to the impoundments Nixon lost unilaterally before SCOTUS than anything so far.
Yes but I think they'd essentially end up creating theories of mind to be successful at it. And it would be a lot harder than they expect. Every problem would lead them to some new problem, and in the end you'd need something like what we'd call AGI.
Great points. If you're an FSD car at an intersection and there are people waiting to cross, you need to understand their intentions as evidenced by their actions, and where they're looking in order to figure out how to navigate. Identifying them as people is not enough.
And formally provably so because to drive you need theory of mind. You can indeed show with maths the limits of maths since it's a formal engineered system with knowable limits, see Gödel etc. We created trains which ban humans from tracks so the system can be safely automated, not so for roads. 1/n
This is a great essay, but I think it fails to point out that it might be (of course we don't know) harder to maintain the epistemic vigilance style, especially when pressed for time. There is amplification of one's style, but perhaps also a gravitational pull away from vigilance.
Yes, all the carriage horses were set free to roam in the woods. It was great for them.
But also, the point of academia IS the work.
He was saying the opposite of that.
Tonight's a good night for every new org to confirm that the DOJ covered up accusations of child rape against the man currently speaking.
Prophetic thread by @benfineman.bsky.social
bsky.app/profile/benf...
BREAKING
Former ICE agent:
On my first day training new cadets, I received secretive orders to teach them to violate the Constitution by entering homes without a judicial warrant.
And then later it was compressed even further to "Harm"
Screenshot from X that reads "Nothing humbles you like telling your OpenClaw “confirm before acting” and watching it speedrun deleting your inbox. I couldn’t stop it from my phone. I had to RUN to my Mac mini like I was defusing a bomb." And then as a screenshot of interactions with some agent that is typing commands that will delete her inbox.
We need a new version of I, Robot where a robot behaves normally for a few days and then suddenly starts killing everyone because its context window got compressed, which changed the first law to: "A robot shall harm a human being or allow a human to be harmed"
“I’m just saying … bsky.app/profile/ronf...
Let's try this again 👋
Hi Bluesky - I'm Massachusetts' Attorney General. You might know me from suing President Trump nearly 50 times, beating Uber and Lyft in court, or being the first woman of color elected to statewide office in MA.
I officially left X today - help me find my MA people?
I think it's better to not even engage in this line of thought at all, because it attempts to put a value on human life according to its learning efficiency, which most of us can agree is just not even on the table.
The inevitable conclusion of Sam's mental perambulation here is straight up eugenics, which is why people who are actually smart know better than to even think in these terms.
These arguments are transparently stupid to anyone who understands that life has value independently of its "efficiency"
