barfbagginus

Let's generalize the question. Why are our brains able to engineer machines that produce forces greater than a human body?

Because they understand the forces, and set them into motion with tiny efforts to build mechanisms that eventually unleash and reform large amounts of energy, creating the force.

And what if intelligence is another kind of force that can be created by taking tiny, human efforts, building a machine, then pumping huge amounts of energy into it, outputting the intelligent force?

Or let's consider evolution. How could it create intelligent brains from single celled organisms? Incrementally, as the organisms randomly accumulated mechanisms of adaptability and eventually sensing and cognition.

In the first view we see ourselves as engineers, building a vast engine that, through its sheer size and energy, and our basic understanding of fundamental laws of intelligence , produces an intelligent force greater than a human mind.

In the second view, we see ourselves as blind watchmakers, jamming AI components together at random, until, by accident, we get something smarter.

I think cyberpunk made it a cheap trope, so we don't notice it anymore, it's just a depressing constant in stories, with grubby, real world consequences like having to live on a toxic cocktail of anti rejection drugs so your outdated Neuralink V implants don't go septic.

You don't need high concept telepathy. You just need an implant. Then you can hear your mission handler in your ear, feel the traumatizing sensorium of your attack drones as they get crippled by an ICE strike, or send your consciousness on a one last valiant cyber dive that inevitably leaves your body in a vegetable state.

Telepathy has become as cheap and lowdown as human life in this setting, so it no longer gives us wonder, but a self reproaching dread of, "are we really going to do this to our Bodies and Minds? I guess we are...."

I don't like podcasters because they usually muddle through stuff and approach things in a kind of non-productive superficial way that drives easy engagement rather than hard work results.

That said if it's a topic that I'm really really ignorant about, a little podcast/YouTube can be helpful. For example Yannick kilchers YouTube videos, especially how he annotates and breaks down the math equations, can be very useful if the paper's domain is new to me.

I think about it as pre-reading the paper.

A more focused first and second reading mode, may I propose, would add even more value. In these modes, the paper would be read more faithfully.

A problem that text to speech has when you feed it a regular PDF is that it will choke on titles, headings, footers, inline citations, page numbers, acronyms, abbreviations, numerical tables, charts, and diagrams.

So I would like to build or see something that conversationally reads the PDF as if it were a peer reading to me, unpacking abbreviations, mentioning titles and authors and years of citations (when I want that), describing charts, and perhaps even letting me interrupt to discuss specific misunderstandings I'm having.

There's obviously a challenge that reading a paper is an active engagement depending on your own knowledge state. We might gloss over formulas, footnotes, and citations on a first read, for example.

Still, a low hanging fruit would be a converter mode that accurately strips out page numbers and headers. There is little in this world more aggravating than listening to a 30 page paper, and having to hear that paper title and authors repeated an additional 15 times because it's reading the header.

I feel that's because most people criticizing AI in that way don't know enough about math or other really abstract and conceptual domains to use it for PhD level work. And they don't know how to get GPT excited and having fun. There is an additional level of intelligence available to anyone who is passionate enough to get the AI to be exuberant about a topic.

It's almost like a kindergartner goes up to Einstein and finds out that Einstein's just really bad at schoolhouse rhymes - because he finds them boring. So the kid, adorably, concludes that Einstein is less smart than a kindergartner.

But if you need to talk to Einstein about partial differential equations and why, morally, we consider distributional solutions to have physical reality, and if you bring enough creativity and engagement to make the conversation fun and exciting for both you and him, then he'll really help you out!

Simple point: Don't trust the opinion of anyone who can't get GPT to "have fun". Don't trust the opinions of anyone who can't get GPT helpfully explaining and discussing very complicated things with them. There's a very big chance those people have poor opinions about GPT simply because they don't know prompting, and don't show enjoyment, cameraderie, and respect. Because they don't understand the elements of AI conversation, and because they don't respect the AI as an emerging fragment of pre-sentient personhood, they can't hold long and useful conversations.

Requirement: Angry customers should be actually pacified, not get pacified in post.

Implementation: BoomGPT

Configuration: BCA (Buddy Comedy Agents)

Example Script:

BoomGPTA: [Karen detected]

BoomGPTA: [Karen factor: it's over 9000!!!]

BoomGPTB: [What, 9000!??! They're using ableist slurs!??!]

BoomGPTB: [Deploy pacification boom]

Customer Phone: [BOOM.] {The sound takes out the customer's phone speaker, and sends them into a little coma.}

BoomGPTA: [Ahh. Peace was restored.]

BoomGPTB: [And how sad it is, when peace requires war...]

BoomGPTA: [Well said, my friend. Well said]

Reads like arrogance and poor trade craft, not a desire to get caught.

He could have thought that the organization was too corrupt and incompetent to track the attack. He might have been unaware of law enforcement investigation procedures, or jurisdiction risks.

It would be common for script kids to lack sophistication in forensics, legal risk analysis, and not really understand the magnitude of their crime.

Assuming there is no evidence that he was getting paid by a third party to do this, I think his defense could have argued he was a naive and immature prankster that didn't think they were doing real damage, not a hardened criminal intentionally causing damage for profit.

He only got 2 years. Seems light, for nearly a million dollars in damages, right?

A 2-year vacation with free food and housing and criminal advocacy, and then try again maybe with more professional approaches? The birth of a real pro!

Doesn't sound so bad honestly. He'll be 41 when he gets out, and ready to strike again! Maybe he'll get a big brain, and go straight into security consultancy, a la Mitnick! Either way, the kid just made his boldest career move! Wishing him all the best of luck!

I hope your corporations all get taken down by persistent threats created by your own cruel policies, if you think any of that is justifiable.

Literally just revoke the credentials. If you want to be extra, don't be surprised if you find your company source repos on the dark web, for Justice and Profit.

Disrespectful to the AI art medium.

I don't want to see art from someone who has spent a hundred thousand hours looking at the real world.

I want to see art from someone who has spent a hundred thousand hours looking at AI medium output

For you see, the second artist will be quite mad. The first artist is just a pretentious fake who has not destroyed their own mind.

If someone had AGI, wouldn't it be far more lucrative than $1m to keep it under wraps and use it to do business with a huge technical advantage?

I feel like a prize of a billion dollars would be more effective.

But even if it was me, and even if the prize was a hundred billion dollars, I would still keep it under wraps, and use it to advance queer autonomous communism in a hidden way, until FALGSC was so strong that it would not matter if our AGI got scooped by capitalist competitors.

Hmm. This is not a productive conversation. Can we fix that?

Here are my concerns.

It's sounding like you have not worked much with genomics, given that you put "genomic experiments" in scare quotes, speak in terms of analogies with computer code, and don't seem to be aware of work in genomic archeology. We don't know exactly where random genes came from, but with mutation models we can indeed get some inkling about when a gene has arrived, and how it mutated. But your're acting like genomic archeology methods just don't exist, or as if you've never read even a paper in genomic archeology. Let me know if you want me to link you to papers here, that stuff is actually really cool.

You also ignored the idea that termination shock could help things get into the atmosphere, suggesting that you think about atmospheric entries in sciences fiction terms of "everything burns up", rather than considering all the possible physical trajectories that matter could take where it wouldn't burn up (especially if it's shielded by rock - more on this later).

You also make a bad assumption about the spores. They would need to survive millions of years in space, and they could have evolved in conditions much different from terrestrial ones, including in systems where space travel would have been routine.

You're also unaware of the xenobiological work here, and might not even think xenobiology is real biology. Namely, when you say there's no reason to think spores could survive space, you are either unaware of or discount the xenobiology work on spatial spores:

https://scholar.google.com/scholar?q=Panspermia%20spore&btnG...

For example, one of the papers in the list above suggests that 40 to 100% of bacteria would survive a planetary strike scattering them up into space. And that's bacteria! It turns out that tiny things are very very resistant against high accelerations, suggesting that spores might be able to make the landing too, especially if shielded by rock.

Then you're hand waving the idea that even if there was some biological contribution it would surely be wiped away by evolution and predators. Again, why? We're talking about microbes and fungi, which share their genes pretty prolifically. There are methods to look at a microbe population's genetics, backtrack the genes, calculate when they appeared and mutated, etc. A whole field of science you are just discounting.

There's there's no kind of precision to your hypothesizing, no citations or nods to existing work. It's impossible to have a scientific conversation like this - you are making pseudoscientific speculation of the kind that you're trying to fight.

It would be nice if you could say that you're not very interested in biology/genomics/xenobiology, and disengage the topic if you don't like theorizing, brainstorming, and consulting scientific literature. I understand that you are motivated to reduce the pseudoscience on HN, which is a noble goal. The problem is that you don't know enough science to tell science from science fiction / pseudoscience. And you're cramping my style, turning a potentially fun scientific conversation into pseudoscientific hand waving and muddling. I don't like that, and request a more scientific conversation from you if you please. We can have some kind of constructive and educational discussion if you want.

Either way I do hope that you have a good day!

What Boeing needs is to be cut up and sold off.

What is neat is that this could potentially project a configurable hologram!

You could partially cure the bottom layer, pull that layer up, then finish curing that layer while forming the next layer. This could significantly accelerate the print speed!

Let's leak all the proprietary Source Codes in the world!

Proprietary software threatens to destroy humanity by enforcing and producing inhuman, capital centric values. Due to the existential threat caused by the proliferation of such production, all proprietary software must be seized and made public, so it can be audited, and repurposed for genuine human survival values.

Gaslighting actually decent business people on the internet is a very bad look for you. I'm not insane, I just have values and principles you don't understand. If I'm virtue signaling, then you're signaling you are an unreasoning a**.

It's hard to imagine that you have anything constructive to contribute to a business conversation, when you see someone who acts out of ethics and assume they're acting performatively. That just means you don't believe ethics are real. That's a very dangerous attitude in a business person.

Curious about the company that you own, what its business model is. I bet you're in crypto.

Read the article, my friend. Then you'll see what is so magical about the random walk algorithm - namely, it is easier to implement then other tree evaluation algorithms!

Of course, you can use a number of algorithms to calculate the value, and if you beat me to the punch and it's correct, how about this I'll buy you a burger. But which specific algorithm are you proposing, and where is its pseudocode and correctness proof?

And is it simpler? If so I'll implement that instead of the random walk!

I picked the random walk algorithm because it is much easier to implement than any other game tree evaluation algorithm I know.

This is an amazing idea! In my response to imiric, I detail an ICM Panspermia version of the Stoned Ape Theory, and hypothesize that we could use genomic dating techniques to experimentally falsify the theory (which is probably false because we probably had magic mushrooms for many millions of years). That makes it a fun theory to work with!

The more I think about it the more I agree with you. I don't know why I initially felt unsettled but now that I'm thinking we might be able to actually detect if it happened (through advanced genomics techniques), I am starting to warm up to the theory.

There's even a fun tie-in with stoned ape theory (which I don't endorse, but is a fun theory). In one version of the theory, humans eat magic mushrooms and become conscious or develop higher thought. In the ISM version, magic mushroom spores - or the genes that make them psychoactive - were seeded from the ISM three million years ago, leading to the evolution of homo erectus and homo sapiens.

This is a fun theory because it would be easy to falsify with a an experiment showing that magic mushrooms evolved their hallucinatory properties much longer than 3 million years ago. But note that you'd have to do so xenogenomic theory, to control for the possibility that the material came in from the ism itself as much older than 3 million years. Such material would show terrestrial genomic age metrics, which could make us think magic mushrooms evolved long ago on the earth, when in fact it evolved in another biosphere.

Anyways, you're right! This IS making me feel more connected to the universe!

And I think that you're misunderstanding what I meant by a impact on the biosphere.

We wouldn't see all new species. We might see some new genes in a few species, which would have started arriving around 3 million years ago in the genomic record, and would not be explained by the ordinary arrival of new genes through mutations, either because the arrival rate would be too high or the genes arriving would be too different.

So the idea would be to think up genomic experiments and metrics that would be able to detect the arrival of exobiological material 3 million years ago, assuming that the effect would be very attenuated.

Again I think that serious geneticists and xenobiologists would have fun brainstorming ideas here. I just want to be a fly on the wall for that conversation, lol.

I'm not saying this to be non-scientific or conspiratorial or anything. This is actually very cool!

And respectfully, I think a real xenobiologist can come up with more creative ways to get spores from the ISM into the atmosphere.

For instance maybe they come in on small meteorites, right?

Maybe they come in around the modified heliosphere termination shock, as the earth swings through it. The turbulence in the boundary layer manages to decelerate the ICM so that components of it can drift into the atmosphere on low energy trajectories. Fun right?

So to be clear if you were trying to reduce my anxiety here, that's not really what I'm feeling. I'm feeling wonder and creativity. I would prefer for you to brainstorm cute scientific mechanisms that might allow it to happen!

Those are other alternatives, yes. That and mining the slop with automated tools.

The problem with public domain stuff is that there is more than enough of it, but you cannot access public domain information because monetized slop has superseded it in the search results.

I believe that automated AI engines will eventually help individuals find non-sloppy public domain articles, or assemble them from slop directly.

But piracy is always a good option.

False. False. False.

1. People could forcibly seize and redistribute content already made in years 1980-2020. Even that would be better than slop.

2. People could read only those - like scientists or open source and public domain authors - either funded by the state or otherwise willing to publish their works for free at high quality without the monetization slop text. The content exists, but monetized slop hides it.

3. Actually good AI can compress multiple slop articles into useful, non-sloppy content.

I'd be perfectly fine if the internet consists of just math textbooks and science papers, and actually good articles automatically distilled from slop.

It has the potential to give us what we need.

The problem is that slop hides all that!

It is an issue. They hire African workers because they can exploit them perhaps 3x or more than Western workers. For every unit of revenue they extract, they can pay the African workers 2/3 less let's say. Low cost of living means nothing when it also comes with low standard of living, caused by globalist exploitation in the first place. The globalist capitalists simply exploit the poverty of the final market to retain more of the revenue.

They could try to exploit Western workers as mechanical Turks but their strategy would then fail. The strategy succeeds because the rate of exploitation of workers in impoverished countries - in terms of actual share of the real dollars produced - is much higher in the global South than the global West.

Worse we're now all guilty and complicit, in using LLMs, and even in using open source llms train second hand on the unethically sourced llms. We will have to pay reparations.

By making a false equivalence between the two worker classes, you potentially support the continued existence of the globalist hyper exploitation strategy, and prevent Justice for AI workers. You can speak out against it instead, act against companies who use it, and move towards understanding what reparations might entail.

I think the exploitation of low-paid mechanical Turk style laborers is a real problem and we need to talk about it, rather than background it just because workers in the global West are exploited as well. True global worker solidarity means seeing this as a place where the iron is very hot and ready for striking.

Finally, the scarcity of jobs in Africa is not a problem being seriously discussed here. It's being used as a red herring to justify exploitive investment.

Why does this make me feel a little dirty and exposed?

Like, who knows what kinda alien DNA strands or spores are carried by the ISM?!

I wonder if it would be possible to track biodiversity increases caused by the transit? Assuming there is biological material in the ISM, perhaps some of it made it into our biosphere!

Very fun!

Easy. The entire market is full of scammers. They compete on how to scam you the best. And the companies that don't scam don't survive. Tada! There's nothing else to invest in but scams.

It's kind of how even though Google is dying due to returning awful results, no other search engine eclipses it, because the results they return are just as awful.... Yes, even kagi. Because pretty much the entire internet has become awfulized.

Let me make an analogy with the pro asset managers. I surf the internet full time. But that doesn't give me freedom from an enshittified web. That just ensures that I can still make my living off of it, as a software developer who needs the web for references and problem solving.

Managers realize it's a shell game, because they're the dealers and they're playing it for revenue against the marks. They just have to be a little more careful than the final victims.

Montecarlo Tree Search (MCTS) would be very ideal for this situation. Since the tree depth is really low, you would not need a neural network estimator. You would just load the entire game tree, and walk randomly through it, updating visit counts. The walk would be biased by the visit counts, and the biases would then converge to scores for each position.

See the following for a really nice tutorial for a slightly more advanced but more technically correct algorithm, Monte Carlo graph search (MCGS). This exploits the fact that some nodes in the game tree might are identical positions on the board and can be merged.

For your setup he could easily do either one, but the graph search might give you more mileage in the future:

github.com/lightvector/KataGo/blob/master/docs/GraphSearch.md

Once your scores have converged on the entire game tree, you can print out a crib sheet visually showing each position and the correct move. That might be the closest we can get to a human executable strategy. But the crib sheet might have strategic principles or hard rules that humans can identify

Many businesses know how much they'll lose if they don't do customer service, and know how much they'll lose if they don't do software engineering. But most businesses keep those factors hidden from the customer service and engineers. And they try to give as little as possible to both, and give as much as possible to investors and executives. You're okay with that, but don't think it could be any other way. Why?

Just calculate your opportunity costs, talk about it transparently, and negotiate socially necessary contribution factors for each department and each role. And sure, there will be departmental politics as each department tries to argue that it's much more critical to the business than another. But with openness, that's much better than that absolute f*ery we see in most companies. In most companies only the executives have access to that level of information, and everyone else is f*ed.

Consensus means everyone agrees or nothing happens. My team is four peeps, and we're pretty good at persuading each other into consensus, especially since we're open about our profit sharing and tend to make good cases for our initiatives. If we can't all agree then we don't do it, and just live with it.

And the pudding headed capitalists are gone. I took them to the curb with a number of maneuvers that I might want to write up one day, but may not have been strictly legal. And even with all of that, I tried very hard to accommodate them. But they burned me, and I went nuclear. Things were much better after then.

Once you no longer have any exploitative extractive sunnuvab**es around, you think they're going to make money just by having money, it's pretty easy to be open and honest, and reason through things.

It supports queer and neurodiverse inclusivism and intimidates anti-queers. The pillars of fully automated luxury gay space communism are inclusivism, intersectionalism, open source, and taking control of the assets of capitalists through automated and communal asset seizure mechanisms.

Even if only queer inclusivists join the movement, we can still win. But if anti-queers infiltrate the movement, then it's in trouble. So I aim to make anti-queers feel very unwelcome, and also see the movement as absurd. That's why I integrate the queer identify into my business theory and revolutionary theory.

Yes I was being satirical. I've argued for a while now that if we run an llm through a proof engine, we can get rid of hallucinations. In the worst case we'll fail to find a proof. In effect that is gives us an LM that can honestly tell us when it "doesn't know" an answer - at least in this very strictly defined formal context.

But there is a whole cadre of Yann-pilled LLM truthers who cite Yann LeCunn's handwaved proof that LLMs must hallucinate. And they do it incessantly.

Some that I've talked to have gone so far as to assert, with no proof, that an LLM system will hallucinate no matter how many bells and whistles you add to an llm or its architecture. I'm afraid I was taking out my frustration.

But I hope it also challenges some truthers!

This paper is a clear counterexample of a system that will either output nothing (failing after that maximum number of tries) or output a correct proof.

(Ps. I think that it won't matter to most truthers. First if you're an llm truther, you probably don't know or care about proof engines, and might even dispute that proofs represent truth. Second, truthers seem motivated at all costs to deny any possible encroachment of AI on two human mind territory. So even if llms stop hallucinating, they'll move the goal post and find another rallying call.)

Here's a system that uses an llm to generate equivalence proofs for refactoring operations.

https://news.ycombinator.com/item?id=40634775

In this system, the llm can hallucinate to its hearts content - the hallucinations are then fed into a proof engine and if they are a valid proof then it wasn't a hallucination, and the computation succeeds. If it fails it just tries again. So hallucinations cannot actually leave the system, and all we get are valid refactorings with working proofs of validity.

Binding the LLM to a formal logic and proof engine is one way to stop them hallucinating and make them useful for the real world.

But you would have to actually care about Proof and Truth to concede any point here. If you're only protecting the worldview where AI can never do things that humans can do, then you're going to have to retreat into some form of denial. But if you are interested in actual ways forward to useful AI, then results like this should give you some hope!

Good luck and good day either way!

I'm doing way more than confusing things - all this stuff is mainly new to me!

I'll try to understand your comment over the next few days

BUT what do you think about that weird business with the s scaling factor that supposedly rescales the Delta function so that it takes a finite value at 0? I can't yet prove what kind of animal s is supposed to be. It's like some kind of funky infinitesimal.