goldfish crackers ©
@sqc.bsky.social
Mostly lurking, in accordance with the robustness principle. In this house we believe distortionary and redistributive issues should be separated, the notion of a "fiduciary" is essential to the future of democracy, and fascism must be destroyed.
Important caveat:
November 6, 2025 at 6:17 AM
Important caveat:
For the non-philosophy-nerds, the "naturalistic fallacy" is an idea in metaethics. It's got essentially nothing to do with the discussion about LLM cognitive capacities or mechanistic interpretability or whatever.
October 6, 2025 at 1:00 AM
For the non-philosophy-nerds, the "naturalistic fallacy" is an idea in metaethics. It's got essentially nothing to do with the discussion about LLM cognitive capacities or mechanistic interpretability or whatever.
From "What do libertarians and pedophiles have in common?" by Joseph Heath.
October 2, 2025 at 6:08 PM
From "What do libertarians and pedophiles have in common?" by Joseph Heath.
Whether or not you believe what Anthropic is saying, LLMs are pretty clearly going to have a huge impact on vulnerability discovery/exploitation.
www.cybersecuritydive.com/news/darpa-a...
www.cybersecuritydive.com/news/darpa-a...
August 27, 2025 at 5:59 PM
Whether or not you believe what Anthropic is saying, LLMs are pretty clearly going to have a huge impact on vulnerability discovery/exploitation.
www.cybersecuritydive.com/news/darpa-a...
www.cybersecuritydive.com/news/darpa-a...
Sorry, but I don't think this is really being fair to physicalism. A physicalist like David Papineau isn't an eliminativist or epiphenomenalist. They believe experience exists and plays a causal role, they just think that experience turns out to be a certain kind of physical structure.
July 25, 2025 at 8:03 PM
Sorry, but I don't think this is really being fair to physicalism. A physicalist like David Papineau isn't an eliminativist or epiphenomenalist. They believe experience exists and plays a causal role, they just think that experience turns out to be a certain kind of physical structure.
"It doesn't mean anything. It's just appears to mean things"
April 28, 2025 at 6:27 AM
"It doesn't mean anything. It's just appears to mean things"
That seems plausible, although Orwell's point about the economics of weapon systems production in "You and the Atom Bomb" seems important here. Early indications from Ukraine point to autonomous weapons being easier to build than we might have thought. Will Hugging Face be the 21st C. AK-47?
April 16, 2025 at 6:59 PM
That seems plausible, although Orwell's point about the economics of weapon systems production in "You and the Atom Bomb" seems important here. Early indications from Ukraine point to autonomous weapons being easier to build than we might have thought. Will Hugging Face be the 21st C. AK-47?
The reason American philosophy departments look so different from European ones is that the Europeans forced out all of their jewish philosophers, many of whom got jobs in the U.S. The dominant paradigm in U.S departments (analytic phil) was basically a German/Austrian invention.
April 8, 2025 at 7:21 PM
The reason American philosophy departments look so different from European ones is that the Europeans forced out all of their jewish philosophers, many of whom got jobs in the U.S. The dominant paradigm in U.S departments (analytic phil) was basically a German/Austrian invention.
I just pasted a bunch of these around my local Tesla dealership. I encourage you all to do the same.
January 21, 2025 at 5:23 AM
I just pasted a bunch of these around my local Tesla dealership. I encourage you all to do the same.
I turned it off 30 seconds in. Imagine saying this to Zelenskyy of all people.
January 6, 2025 at 2:15 AM
I turned it off 30 seconds in. Imagine saying this to Zelenskyy of all people.
December 23, 2024 at 4:32 AM
That's how I read this exchange.
December 23, 2024 at 12:54 AM
That's how I read this exchange.
December 22, 2024 at 11:23 PM
Let me try to convince you then. The underlying issue is this: if we want to explain some cognitive capacity of ours (say, vision), we can't do it by appealing to an external observer. There isn't a little person in your head, looking out your eyes. What vision has to be is something sub-personal.
December 21, 2024 at 8:50 AM
Let me try to convince you then. The underlying issue is this: if we want to explain some cognitive capacity of ours (say, vision), we can't do it by appealing to an external observer. There isn't a little person in your head, looking out your eyes. What vision has to be is something sub-personal.
Explain to me how this is done with just the reductionist vocabulary you use to characterize LLMs in general. If there's no emergence, no cognitive level of description here, how is this system able to produce these outputs?
December 21, 2024 at 7:28 AM
Explain to me how this is done with just the reductionist vocabulary you use to characterize LLMs in general. If there's no emergence, no cognitive level of description here, how is this system able to produce these outputs?
December 21, 2024 at 7:01 AM
This is the context. I was responding to someone else, and it was relevant *to that conversation*. You don't have to care about fluency if you don't want to.
December 21, 2024 at 5:52 AM
This is the context. I was responding to someone else, and it was relevant *to that conversation*. You don't have to care about fluency if you don't want to.
Fwiw, the two papers that got me thinking "oh shit maybe these artificial neural nets are good models of *our* kind of cognition" were:
(1) pmc.ncbi.nlm.nih.gov/articles/PMC...
(2) www.pnas.org/doi/pdf/10.1...
Both use ANNs as suprisingly good models of different brain areas. Mind-blowing imo.
(1) pmc.ncbi.nlm.nih.gov/articles/PMC...
(2) www.pnas.org/doi/pdf/10.1...
Both use ANNs as suprisingly good models of different brain areas. Mind-blowing imo.
December 21, 2024 at 4:45 AM
Fwiw, the two papers that got me thinking "oh shit maybe these artificial neural nets are good models of *our* kind of cognition" were:
(1) pmc.ncbi.nlm.nih.gov/articles/PMC...
(2) www.pnas.org/doi/pdf/10.1...
Both use ANNs as suprisingly good models of different brain areas. Mind-blowing imo.
(1) pmc.ncbi.nlm.nih.gov/articles/PMC...
(2) www.pnas.org/doi/pdf/10.1...
Both use ANNs as suprisingly good models of different brain areas. Mind-blowing imo.