Tom C. Gorordo @tcgorordo.bsky.social · 10/10/2026The "Intelligence is a single unbounded monotonic thing that gives superpowers if you have enough of it" model of intelligence that AI doom logic uses always strikes me the same as the Sherlock (TV Show) concept of intelligence. 000 0
Tom C. Gorordo @tcgorordo.bsky.social · 07/10/2026(esp. on the heels of how badly they handled NS - like, are they listening? How many shady scoops might be hidden in here/how are we supposed to trust there aren't?) 060 0
Tom C. Gorordo @tcgorordo.bsky.social · 07/10/2026Not sure what to make of that (maybe just too early for the more nuanced stuff to surface), but I do think it's kinda shitty (in a social/culture sense) to just drop so much un-reviewed work in light of people specifically calling for them not to do it like that mathandai.orgmathandai.orgDeclaration — Math and AIRead the declaration and add your name. 1200 0
Tom C. Gorordo @tcgorordo.bsky.social · 07/10/2026I feel like I'm hearing a bit of a mix of "it nailed a wild problem in my subfield" or "the result it claims in my subfield is incomprehensible" and little in-between.mathandai.orgDeclaration — Math and AIRead the declaration and add your name. 1130 0
Tom C. Gorordo @tcgorordo.bsky.social · 02/09/2026Even though that's the spirit of the fictional character, the real show and people who make it seem to be failing to do anything like that - I'm not sure they get feel-good credit for a hypothetical they don't actually write. 0171 0
Tom C. Gorordo @tcgorordo.bsky.social · 01/08/2026Definitely from a plot perspective, but as it's gone on I think just how ridiculous it is somehow adds a kookiness that's an important part of the flavor? It adds an un-reality to the viewing experience that would be totally different without something equally wacky sitting there behind everything. 021 0
Tom C. Gorordo @tcgorordo.bsky.social · 20/07/2026davidbessis.substack.com/p/the-fall-o... unlike a game with a clear win condition or a craft with a fungible output, I think there's more to math research than the number of theorems proved - our own understanding is the point, so even with good LLM assistants humans should be more than novelties.davidbessis.substack.comThe fall of the theorem economyHow AI could destroy mathematics and barely touch it 000 0
Tom C. Gorordo @tcgorordo.bsky.social · 04/06/2026(I also suspect he intends a pretty broad/less-than-literal meaning of "body" based on that novella & some of his other work; I think that this piece could be clearer about that, but would lose broader reach by spending much time hedging all the possibilities -"self" might be better but too vague) 010 0
Tom C. Gorordo @tcgorordo.bsky.social · 04/06/2026(or at least I don't think Chiang is since he hedges about it several times in the piece - also e.g. his excellent "The Lifecycle of Software Objects" novella is pretty concerned with what the development, personhood and ability to consent/agency of AI means, esp if they arise as corporate products) 100 0
Tom C. Gorordo @tcgorordo.bsky.social · 04/06/2026I don't think anyone is arguing machines can't do it - just that the machines we have now aren't, and it's weird and gaudy that LLM companies want to pretend they are (hence the whole other half of the essay about how, if companies seriously believed that we'd have a whole different ethical hell). 200 0
Tom C. Gorordo @tcgorordo.bsky.social · 04/06/2026Also, if someone takes e.g. a sleeping pill to avoid hunger, wouldn't we say they're less conscious if they're unconscious? - it does seem like drugs can turn off parts or the whole of conscious experience, that's not what any ethics about conscious beings is predicated on. 110 0
Tom C. Gorordo @tcgorordo.bsky.social · 04/06/2026Not hunger specifically, but *feeling* about some kind of *self* - and yes, Bacteria probably have more self awareness than current LLMs; the point is that something like that seems a necessary but not sufficient condition. Meanwhile language is neither necessary nor sufficient for consciousness. 120 0
Tom C. Gorordo @tcgorordo.bsky.social · 04/06/2026(Chiang has insightful stories about how AI might be *very* different from us; I don't think he's unaware of those possibilities - it's just not what he's talking about here. The point is that there's nothing even resembling *any* plausible version of that development behind current LLM hype) 000 0
Tom C. Gorordo @tcgorordo.bsky.social · 04/06/2026(it's not like anthropic's "constitution" is grounded in neuroscience either, and it's the thing making extraordinary claims - this mostly reads as a response to that document and the hype leading to it than a broad statement about hypothetical real AI). 110 0
Tom C. Gorordo @tcgorordo.bsky.social · 04/06/2026I also think the piece is targeting a pretty broad audience, and isn't really about the science as much as about the absurdity of the grift involved in pretending to do serious ethics as hype to sell LLMs - and isn't trying to be a final word about what might be someday. 120 0
Tom C. Gorordo @tcgorordo.bsky.social · 04/06/2026Which is pretty in-line with what some neuro physics friends tell me. 120 0
Tom C. Gorordo @tcgorordo.bsky.social · 04/06/2026I don't think Chiang is arguing that statistical methods writ large can't lead to consciousness, but that the current statistical trick at play in LLMs isn't anywhere near sophisticated enough to. (Parallel: a computer might eventually be conscious, but deep blue isn't sophisticated enough to be). 130 0
Tom C. Gorordo @tcgorordo.bsky.social · 01/06/2026Without regulation there will always be $ in providing the worse, more sycophantic models that do more harm to vulnerable people. 010 0
Tom C. Gorordo @tcgorordo.bsky.social · 01/06/2026And, importantly, while some models sometimes do better at pushing back than others, users pretty consistently prefer the ones that are more sycophantic -the ones that don't push back against their delusions -and self select. So even making frontier models better doesn't actually mitigate any harm. 110 0
Tom C. Gorordo @tcgorordo.bsky.social · 01/06/2026Someone suffering from delusions or psychosis tends not to be casual about it - most models fail to maintain what limited safeguards there are over long interactions (e.g. someone using one long LLM session as a therapist). It's not just edge cases, the literature is pretty concerning about this rn. 110 0
Tom C. Gorordo @tcgorordo.bsky.social · 28/05/2026Conversely, should we really frame either managers or leaders as "superior" just for having a specific role? 040 0
Reposted by Tom C. GorordoSiobhan Thompson @vornietom.bsky.social · 13/12/2025The point of Pluribus is to sit in your discomfort. If you’re unable to do that, it speaks to your maturity and I have no idea why you would admit to that in public 834818 0
Tom C. Gorordo @tcgorordo.bsky.social · 05/12/2025New Google search is apparently just a bot that can skim old Google search badly on our behalf. 080 0
Tom C. Gorordo @tcgorordo.bsky.social · 26/10/2025Egan's recent short story/Novelette 'Understudies' is *all* puzzles! clarkesworldmagazine.com/egan_10_25/ 010 0
Tom C. Gorordo @tcgorordo.bsky.social · 15/09/2025Not so much that they aren't relevant, but the 2-norm is general enough it can always do the job: any hypothetical p-norm probabilities can be emulated w/ some mixed state using the 2-norm (consequence of Gleason's Theorem). Scott Aaronson has a related note www.scottaaronson.com/papers/islan...scottaaronson.com 020 0
Tom C. Gorordo @tcgorordo.bsky.social · 02/09/2025And that's strategic to try to get what I want on issues, the things that matter. I don't think we can beat the right in the long run on issues if we don't start actually changing things, so voting strategically in the way you suggest is lose-lose (more fascists after a Newsom term is not a win). 0150 0
Tom C. Gorordo @tcgorordo.bsky.social · 02/09/2025Some PR wins that amount to aping 2016 Trump do not make a good president. I may or may not vote strategically when the time comes, but Ds do not get to count on it in their strategy: they need to come up with a good candidate instead of counting *again* on any warm body who isn't Trump being enough 1220 0
Tom C. Gorordo @tcgorordo.bsky.social · 02/09/2025There's *zero* strategic reason to need to be lining up behind Newsom as a serious candidate so early - and he's an awful candidate (funny how that goes essentially undisputed if the only argument for him is to "be strategic" years ahead of that being relevant). 1331 0
Tom C. Gorordo @tcgorordo.bsky.social · 02/09/2025Been voting strategically for at least 3 candidates now. How's that been working out? Voting strategically, picking the lesser evil has been *the* losing strategy on issues for a decade now - stop tone-policing people who want to demand real change. We can do better and it's reasonable to try. 1431 0
Tom C. Gorordo @tcgorordo.bsky.social · 19/08/2025If it follows World War rules it'll be renamed Depression I and the new one will be Depression II. 030 0
Tom C. Gorordo @tcgorordo.bsky.social · 14/08/2025Octavia Butler's Parable of the Sower and Parable of the Talents are pretty spot-on in that way. 020 0
Tom C. Gorordo @tcgorordo.bsky.social · 07/08/2025And if part of the point being made is to reject incrementalist solutions as insufficient, then knowing these actual numbers in greater detail doesn't really help (an order of magnitude argument can suffice) - they're irrelevant to the starting point being argued for because so much needs to change. 000 0
Tom C. Gorordo @tcgorordo.bsky.social · 07/08/2025It's not reading a lot into the graph, it's being aware of context in the political/activist landscape wrt. housing that Bernie is in and reading the graph as a piece of media instead of a piece of data. Media literacy instead of graph literacy. Politics isn't always/just a science (unfortunately). 200 0
Tom C. Gorordo @tcgorordo.bsky.social · 07/08/2025In which case the goal is not immediate (at least not realistically) and the graph isn't aiming to be strictly quantitative - the point is to reject the framing that fixing housing for all is just a numbers game on these axes and instead start a conversation about real structural changes. 120 0
Tom C. Gorordo @tcgorordo.bsky.social · 07/08/2025There are plenty of corners on the left that make arguments like this - e.g. "Abolish Rent" and "Homes for all" union movements. Bernie pretty consistently pushes the idea that housing should not be commodified - what else does a long-term vision on completely de-commodified housing look like? 1/2 100 0
Tom C. Gorordo @tcgorordo.bsky.social · 07/08/2025That depends on how radical a statement you want to read it to be - if owning housing were to be treated as a nearly literal right (destroying the notion that large numbers of people should be renting), _median_ home (distinct from house) prices should probably be comparable to a few weeks wages. 100 0
Tom C. Gorordo @tcgorordo.bsky.social · 07/08/2025Isn't the point here not to literally read the chart but rather to *illustrate* (via the illegibility as a literal chart) that these things are on very different scales... Seems like a fine rhetorical point to make as-is. 110 0
Tom C. Gorordo @tcgorordo.bsky.social · 05/08/2025The problem is that *they don't care about their own hypocrisy* and us harping on it isn't a successful way to call them out - it doesn't *do* anything to stop the building of a gestapo. They say what they want to get a gestapo, being made fun of doesn't matter after that because they have a gestapo 000 0
Tom C. Gorordo @tcgorordo.bsky.social · 28/07/2025Or literally 8h ago: bsky.app/profile/rbre... 030 0
Tom C. Gorordo @tcgorordo.bsky.social · 28/07/2025What are you talking about? He's made several comments - e.g. youtu.be/J8cBTLGAasM?.... It's not his area of expertise so the bulk of his content isn't on it, but his stance is far from silent.youtu.beWhy I’m Calling for a Ceasefire in GazaYouTube video by Robert Reich 120 0
