re: Logic Drift

i hate slop. my primary goal for the next 4 quarters is to use minimum 1 billion tokens to build the best second brain slop calendar app of all time

[re: Logic Drift @htmldotyuck]

Wouldn't the answer be to refer to your co-authored Science paper, throw out the idea of a 1-for-1 citation list, and reframe writing in terms of the external and "internal 'societies of thought,' spontaneous cognitive debates that argue, verify, and reconcile to solve complex tasks"? Openly consider how this (for lack of a better term) "hybrid" research is being done. I, at least, would find that useful.

(Otherwise, lol @ the idea of plagiarism here, how would you even try to argue about that; it's like everyone just read the work of art in the age of its mechanical reproduction for the first time and missed both the core argument and how to interpret approx 100 years later...)

[re: Logic Drift @htmldotyuck]

@jaycaspiankang @hiattb I know you just created a burner that likes data centres but isn't it a bit early to start saying there's so many of u fans u need to be run like an organisation? 1 that's not fandom in pure form 2 big "we demand to be taken seriously" energy. wtf man

[re: Logic Drift @htmldotyuck]

This is the explanation in its purest form. The feelings of hopelessness and fear are valid but nothing more than a misidentification of the Sisyphean task of treating psychotic Zionists as reasonable -- if we share enough the boulder will cease to roll back down. The internet is the Arab Spring. If leftists want change they have to get offline and put their lives at risk, far greater risk than campus protests. You don't achieve progress on what Mark Carney calls a New World Order being built in media res by posting.
One video of gore is enough to convince many. For the rest, no quantity will convince: genocidaires exist in a different reality. You are not wrong that to keep a pot at a constant boil and never let it boil over (take action that risks poster lives) is worse than useless. It pushes people to the belief that nothing ever changes, history is predetermined. Rage without external release inevitably turns inwards.
(Finally, who seriously believes a Contrapoints video is going to spring the masses into action, an action they cannot decide on because it's not clear what would work!) You're merely a convenient punching bag because for a brief period your excellent videos genuinely changed some minds. That's not the case anymore, for reasons that have nothing to do with you. As always with social media the option is just to log off sadly.

[Natalie Wynn @ContraPoints]

I’ve been waiting for this to blow over but it’s been years, so I have to address it: can one of the 100,000 leftists on here for whom it’s an essential article of faith that I am a “genocide-loving Zionist” explain—why? Despite my condemning the genocide dozens of times and never saying anything positive about Israel in my entire life?

[re: Logic Drift @htmldotyuck]

another person who loves sports got so pissed off they cancelled outright too? no way
https://t.co/Ja3gSLteXr

[🏳️‍🌈 Spencer Keaton 🧀 @SpencerRose3]

Funny how that goes @Sportsnet

[re: Logic Drift @htmldotyuck]

@Sportsnet @SportsnetPR claim to be inundated with requests today. Wonder why? Maybe its the increase in price from $249/yr to $344/yr between fall 2024 and summer 2026. 'to bring the best streaming experience," like a national blackout of over 10 min during the Jays WS run. borderline impossible to speak to a human, and the streaming tech might even be worse now all while Ed Rogers fills his already bulging pockets. Borderline extortion at this point. time for the @CanadianPM to have another look at competition laws in this country

[re: Logic Drift @htmldotyuck]

Whether bullish or bearish on AI, I don’t see this problem going away. Agree with @ClementDelangue’s ideals but see zero path to that future in a USA that’s barely, if at all democratic (see the Mythos and Fable drama). Maybe if there were a mass (populist?) push to open source, but the hype/doom divide currently makes some sort of unity hard to even imagine. (Note I’m Canadian so these are observations from across the border.)

Friends often come to me to try to assuage AI anxiety; I've had to learn to basically code switch lest I also get accused of psychosis - because to explain to someone who knows nothing but how to use a GUI what's going on is like trying to convince magic is real; it breaks brains if it doesn’t get tuned out. Ultimately I’ve found that turning to analogy that doesn't even reference computation has worked best.

IDK that I have anything to contribute to Antikythera & AgentWorld beyond this tweet beyond a desperate call for open-sourcing. But I'd add one thing: a step I think matters more than many would like. And it's all about narrative.

Humans are anthropocentric and fearful. The Science paper on "centaurs" was great for me but it's the sort of thing that'll also scare the shit out of people. Forget universities, oligarchs, geopolitics. Calling for open source is good but not enough. There needs to be genuine work put towards helping regular folks understand AI in a way that isn't just an eschatology. Pure communication that it's not gonna end their world and feed dystopian sci-fi fears.

I think that's as urgent a task as any actual technological development. Separate the political crap, the idiot C-suites, the doomers and hypers, and dedicate real time and resources to one simple task: explaining wtf is happening. Teach people what to actually fear and what's bullshit. An ongoing task. Maybe boring; certainly not cutting edge - more like the Apple Store sessions teaching seniors to use their phones. Help people understand what we have is a tool but not an apocalypse.

What’s unique (but not really if you know your Phil and Hist of Science and Tech) is the tool is being built without knowing its direction or future (speculative claims for IPO valuation don’t count). But that’s a doozy of a concept to explain, and it’s certainly not helped by endless claims that AGI/ASI is moments away (I’m bearish on that timeline and the whole agents training agents theory of recursive learning towards consciousness or cognition. And even if I’m wrong it scares the shit out of people when they dont know what it means).

Are those late night musings relevant or useful? I have no idea. I just know there's zero attention to the people side of AI beyond C-suite morons claiming you’ll lose your job or vague fear-mongering think pieces that don’t actually understand a single technical detail.

My opinion: it's making things way worse. Im 32 yo. I remember the excitement of dial-up, but too young to recall the advent of the WWW and open-source optimism. I have a hard time seeing a peaceful path forward for AI unless some of that optimism is recaptured from the oligarchs and translated to the masses. /end
https://t.co/itk6SLE4Fp

[clem 🤗 @ClementDelangue]

Concentration of power, capabilities and economic wealth is the biggest risk in AI. We need open science and open-source more than ever!

[re: Logic Drift @htmldotyuck]

In fairness, people are scared, it's not going away, but the biggest problem is they aren't sure what to be afraid of beyond 1 a sense their future is foreclosed and 2 something akin to early fears upon the invention of train travel (forced isolation, an "agoraphobia" where there's no option to step off or away, and public commentary focussed excessively on the possibility of psychosis, as trains historically and AI now alter human perception and experience of time; @bratton's tweet quoted here about the linked and bad article directly addresses that public "commentary").

A lifetime ago (2020) I finished an MA thesis called "Imaging Information," probing at "post-truth" fear mongering and hypothesizing the real problem was not "misinformation" but how going online into an unreliable info env produced a weird inversion of Shannonian info theory - high entropy noise shifted towards associations with sources of trust, because a simple, clear signal was cause for suspicion (if everything online is a chaotic mess, a cogent signal turns into a red flag). Obv the powers that be in STS called bs and did not like that argument because it contradicted their work on “misinformation” (yikes on how that aged) and cushy TT positions.

I wanted to do a PhD delving deeper into the compute element of this hypothesis - how humans don't experience "the stack,” let alone a silicon chip as it functions. Scale and timeframe aren’t legible to a human brain, so it's magic, a story, a "how does electricity produce that" that freaks people out. I planned to start with digital image creation and destruction via future-looking (micro-scale) GPU production of moving images in anticipation of the coming AI boom; needless to say I got told it was all junk.

Now (or for now) we have LLMs: the ultimate vehicle to sign up for, isolate, and tell yourself stories, but more often nightmares (the bulk of press coverage). It doesn't help the marketing strategy for a company like Anthropic is "scare people about powerful models"; add in the geopolitical USA vs China fear-mongering, profiles about creeps like Altman, and I find it hard to blame people for their fear. Especially in Western economies where public knowledge of AI usage that stands out is limited to: stupid vibe-coded get rich schemes (or junk that doesn’t work); isolated youths with chatbot best friends (bad and sometimes tragic); and an economy where you can’t afford RAM or SSDs currently and Anthropic is releasing white papers that suggest they’re high on their own supply, building an ASI God with the tone of Mark Fisher and getting scared of data they hard-coded in when Claude tells them a story (yes, I despise the public face and message of Anthropic). https://t.co/AaBlrjDsEN

[Benjamin Bratton @bratton]

This is far from unique and communicates essentially : "I am professoinal commentariat -- but am lost. AI feels scary and out of control and I want it to not feel scary and out of control, so somebody should do something! That is my big idea. The end."

[re: Logic Drift @htmldotyuck]

Don't optimize your kids. Allow them to explore, use tools, discover intelligence and learn. I don't wanna hop on the anti-AI train but this old ANT stuff is a recipe for disaster and prob wont help ur kids with the future to come either, to be blunt

[re: Logic Drift @htmldotyuck]

The QT is a cool Obsidian-style "second brain" network visual, reminiscent of 30yo actor-network theory so less novel than it seems. & nothing against making math fun, but do we need to optimize kids like token usage? Brains also develop through fun & play. Kids aren't machines

[Rui Ma @ruima]

This is pretty incredible. Free idea, yall should advertise on Xiaohongshu for all the Chinese parents yo. Most importantly, if this is really a more efficient and therefore more fun way of learning math this is actually really positive for the kids

[re: Logic Drift @htmldotyuck]

graphs, network visualizations like this: functionally meaningless. It looks like something but it's a black box. And im the opposite of a Luddite. Humans aren't Obsidian visualizations. Don't always agree with the args, but this @bratton et al i do 1/2 https://t.co/0UMZU96wRi

[ArtButMakeItSports @ArtButSports]

Dance in the Country, by Pierre-Auguste Renoir, 1883 (flipped)

[re: Logic Drift @htmldotyuck]

Sometimes I wonder what it would be like to read Dante describing Virgil guiding him down to the 10th and final circle of Hell, Polymarket. Perhaps I should get a multi-agentic AI to compose this final chapter.

[Polymarket @Polymarket]

BREAKING: Trump declares “I’d be the greatest communist in history.”

[re: Logic Drift @htmldotyuck]

@jfazli @ChelseaGBH Yea the way he kicked the ball away was probably the speed of a bullet he could've killed him. And then the simulation where he entered the matrix I would ban him for life the little disgusting punk. You have a career in broadcasting in you I can tell, the way you think the game

[re: Logic Drift @htmldotyuck]

@jfazli @ChelseaGBH Fairly certain they both got yellows but maybe I have the wrong channel

[re: Logic Drift @htmldotyuck]

@jfazli @ChelseaGBH I feel like xhaka going spikes up in return is a little more likely to seriously to cause injury, no? In any case they both got yellows and ur gonna gonna be ok I promise

[re: Logic Drift @htmldotyuck]

@thecollegehill @Afinetheorem @ezraklein In fairness, guy a is at the most ridiculous business school in Canada, and blunt rotation could equal incredible bearing witness to, or simple gossip about, the sex that invariably happens as the dialog is infused with libations. I'd die by seal team 6 trying to get a good view