Tom Ault

@basedganyu A first date is supposed to be low pressure so you can start getting to know each other without expectations and have an easy exit in case it turns out you're not compatible. Coffee, a movie, etc. are all good options.

"I want to go on a date" Ok where "No you have to plan it, otherwise you're a cuck" Oh ok, well- "Also, we're not allowed to go anywhere or do anything, I have a very specific idea of a perfect date and if you get any aspect of it wrong, I'm ghosting you" ....

@rolandbouman Nobody reviews compiler output because compilers generate code deterministically according to rules contained in a formal language specification against which that output can be verified, giving compilers a level of trust that would be unwise to give to an LLM.

Compiler construction is one of the oldest, best understood CS fields. It's decades of work by the brightest minds, and it's grounded in logic, informed by experience and strictly deterministic. Comparing that with LLM-based coding agents is just wrong. x.com/IceSolst/statu…

@willchen500 The secret to Harvey's success is that they don't just market to law firms, but to the clients of those law firms, who turn around and ask the law firms "How are you using AI to reduce my bills?" and "Why am I paying so much for associates when Harvey can do the same thing?"

Harvey and Legora are essentially sales organisations that resell tokens. They have hired legions of ex big law juniors and mid levels as sales people (“GTM”) along with some ex partners to wine and dine their former colleagues. They slap on a UI that makes them look different from ChatGPT but the product differentiation and vertical specific features are far and few in between. You could just as well use both for any white collar job. Their web apps are basically 1. A chatbot interface 2. A projects function where you can upload your files 3. A tabular review function where you can bulk review documents in a table 4. Workflows which are just custom prompts you write for the chatbot or tabular review. I was able to build everything plus some additional functionality they do not have like version control in mikeoss.com in two weeks. I call this the “token reseller theory”. They are like car dealers or real estate agents but for tokens. The model providers get them to do the selling to crack open the reticent legal market. What happens to H/L now that the model providers want the market for themselves? Does not bode well for them.

@Shiro_Shihi If a Japanese company had built GitHub, you'd have to fax your commits.

日本企業がGitHubを構築した場合はこう

@willchen500 BigLaw is paying for Harvey and Legora because their clients are demanding they "use AI" to drive down billing costs, and Harvey and Legora are names their clients know.

From the messages that I’m getting from Biglaw associates, I am just now learning that some firms in fact ban their lawyers from uploading documents to Harvey’s vault and do not implement Harvey’s and Legora’s agent harnesses to DMS due to security concerns. So no agent, no tabular review. Biglaw is paying 1000+/lawyer/month for an ultra expensive chatbot, chatGPT circa 2024, and they are happy with it.

@wil_da_beast630 Less than 1% of men have an erect penis of eight inches or more, so either she's full of sh*t or close to 100% of women have very unrealistic expectations.

@corsaren - The guy obsessed with the Chinese Room Argument even though it's obvious he doesn't understand thar argument. - The guy just fronting so he can get a spot on Dwarkesh and hype his startup/memecoin.

types of guy in the AI consciousness debate: - guy who thinks ai can’t be conscious because it’s “just a stochastic parrot” - guy who thinks ai must be conscious because claude is a good boi - guy who hasn’t gotten over 4o - guy who unironically thinks everything is computer - guy who claims to have a more nuanced argument for computational functionalism, but it just boils down to everything is computer - dualist whose belief in dualism is downstream of their belief in god, yet tries to argue the inverse - guy who doesn’t understand the difference between cognition and p-consciousness - guy who asserts illusionism but has apparently wrestled with zero of the implications other than “reductive materialism wins again” - guy who says the hard problem is easy, but then proceeds to only answer the easy problem - guy who rejects ai consciousness because otherwise it might be wrong to abuse claude with death threats to make CRUD apps faster - guy who argues that consciousness is is the key to moral patienthood, but completely ignores that when discussing animal rights - eliezer yudkowsky being pedantic - guy being pedantic about eliezer yudkowsky’s pedantry - guy who rejects dualism because that would make mind uploading impossible and mean that he finally has to confront the inevitability of his own death - guy who thinks this argument is unresolvable so everyone should just shut up and accept his position (which obviously deserves the benefit of the doubt) - guy who would literally cut off his own hand if he thought there were a 1 in 10 trillion chance of creating ~infinite utility~ - guy who just thinks that redness is, like, super weird, man. can’t explain that! - guy with a rarely-updated philosophy blog despite not majoring in philosophy or even reading that many books, talking about how “the whole field is up its own ass” - academic philosopher who, for some reason, expects a higher caliber of discussion on x dot com the everything app - guy who thinks that vectors are literally emotions and bites the bullet that, yes, your thermostat does feel hot - panpsychist who took dmt once and contributes almost nothing to the conversation - guy who is literally a solipsist but is still really invested in convincing strangers on the internet that he’s right any that i missed?

@TheNoamLewis @_trish_xD Mocking for the sake of mocking or living up to some testing dogma just brings pain because,all those mocks have to be maintained along with the original class.

@TheNoamLewis @_trish_xD Using mocks when the original class would have worked just fine. Mocking out databases and services that are slow, expensive to stand up or not guaranteed to be available when the test is run is a practical necessity.

I can't remember any "best practice" advice from tech blogs that actually held up. But every principle from old systems papers turned out to be exactly right, and I should have read them sooner. Something went very wrong when software got easy to ship. People stopped understanding what runs underneath. Old constraints teach better than new abstractions.

@TheNoamLewis @_trish_xD Test mocks are the worst. They effectively create more classes or functions you have to maintain along with the originals. Heaven help you if you don't have an easy way to find all of them.

@_trish_xD I don't know about old systems papers, but maintaining a large project is easier if it has less fancy abstractions and just decomposes cleanly into obvious parts. My worst enemy is dependency injection frameworks which introduce magic non-linear flows, and test mocks

@uncledoomer Is the distance Euclidean distance, distance along the surface of the Earth if it's surface were flat or distance along the actual surface of the Earth, including terrain?

this hypothetical brings out two types of people: people who would take this deal in a heartbeat and be 100 million richer within 3-4 weeks, and people who failed geometry

@TheDariaBlues @DungeonNoir Star Control 2 would make a great sci-fi setting. Your campaign could be set in first war against the Ur-Quan or during or after the liberation of the Alliance. Just treat everything from Star Control 3 and after as non-canon.

Question for the elderly (e.g. @DungeonNoir): What are some old PC games (1980s-1990s) that could make for an interesting tabletop RPG setting? I learned about "Syndicate" earlier and it looked fun. "X-COM" is another (although it has a modern reboot).

@CorpseKings The Corpse Brides

What's a cool name for a Coven of Witches that Worship the Undead? 💀

@Grand_DM The 10' pole is why the trap door is 10' away from its trigger.

What piece of mundane equipment do you recommend? I’m going with the 10-foot pole. It’s a metaphor for a style of play where you visualize the environment, ask questions, and use simple tools to explore it. Exploration becomes something you solve, not something you roll for.

@memeslich Fun.

What's something about AD&D you would bring to the current Dungeons & Dragons?

@CorpseKings Forcing the other players to adhere to the increasingly ridiculous and bizarre tenets of your faith (that you just made up) in order to receive those sweet, sweet cure spells.

What's the best part of playing a Cleric in D&D? 🤔

@peach2k2 As stated here, the problem is underspecified. If the predictor (P) is adversarial, it always predicts both boxes and you get $1K. If P is infallible, you pick box B and get $1M. If P is highly accurate but fallible, the problem is much harder and you can argue for both choices.

been thinking for an hour on this but if you go for both boxes you're geniunely subhuman

@SuryaGanguli @aran_nayebi @Leokoz8 @michaelandregg @oh_that_hat @eonsys @alexwg @Philip_Shiu @AdamMarblestone Science is secondary to fundraising. Cool demos and overhyped claims drive the latter.

@aran_nayebi @Leokoz8 @michaelandregg @oh_that_hat @eonsys @alexwg @Philip_Shiu @AdamMarblestone It’s quite irresponsible to post such videos without explaining answers to fundamental questions like this. This not how science should be communicated.

Does anyone know how this virtual fly moves *without* RL, given that the actual motor neurons weren't traced out (because the body wasn't scanned)? @Leokoz8 @michaelandregg @oh_that_hat @eonsys @alexwg @Philip_Shiu @AdamMarblestone

@LavenderGhast The party is effectively the old royal court and the chairman is effectively the emperor.

China is a funny case because they adopted communism, then charged around trying to destroy all of their own history so only current communist glory exists. Only to then, decades later without openly admitting it quietly drop communism for capitalism and re-adopt all their ancient cultural heritage. Only keeping one aspect, which is the party can do whatever it wants. It's like bragging about moving into a state of the art future house to all your neighbours. Moving in to find its actually shoddily constructed and falling apart. Then quietly moving back into your old place and only keeping the couch.

@vikhyatk 5x as many RL environments for 5x as long using the latest optimizer and verifiable reward. To a founder. a world model is what I'm hyping as my moat to get VC's to cough up more $$$$. To a VC, a world model is the thing I'm afraid I'm missing out on.

@vikhyatk To a sane professional, a world model is a model of the environment useful for explaining outcomes and predicting the result of actions. To an AI Researcher, a world model is what I'm hoping to get by training my model on another few petabytes of data, post-training it on...

not sure what a world model is and at this point i’m too afraid to ask