> With the emergence of AI in science, we are witnessing the prelude to a curious inversion – our human ability to instrumentally control nature is beginning to outpace human understanding of nature, and in some instances, appears possible without understanding at all.
A while ago I read "Against Method" by Paul Feyerabend and there's a section that really stuck with me, where he talks about the "myth" of Galileo. His point is that Galileo serves as sort of the mythological prototype of a scientist, and that by picking at the loose ends of the myth one can identify some contradictory elements of the popular conception of "scientific method". One of his main points of contention is Galileo's faith in the telescope, his novel implementation of bleeding edge optics technology. Feyerebend argues that Galileo invented the telescope as primarily a military invention, it revolutionized the capabilities of artillery guns (and specifically naval artillary). Having secured his finances with some wealthy patrons, he then began to hunt for nobler uses of his new tool, and landed on astronomy.
Feyerabend's point (and what I'm slowly working up to) is that applying this new (and untested) military tool to what was a very ancient and venerable domain of inquiry was actually kind of scandalous. Up until that point all human knowledge of astronomy had been generated by direct observation of the phenomenon; by introducing this new tool between the human and the stars Galileo was creating a layer of separation which had never been there before, and this was the source of much of the contemporary controversy that led to his original censure. It was one thing to base your cosmology on what could be detected by the human eye, but it seemed very "wrong" (especially to the church) to insert an unfeeling lump of metal and glass into what had before been a very "pure" interaction, which was totally comprehensible to the typical educated human.
I feel like this article is expressing a very similar fear, and I furthermore think that it's kind of "missing the point" in the same way. Human comprehension is frequently augmented by technologly; no human can truly "understand" a gravitational wave experientially. At best we understand the n-th order 'signs' that the phenomenon imprints on the tools we construct. I'd argue that LLMs play a similar role in their application in math, for example. It's about widening our sensor array, more than it is delegating the knowledge work to a robot apprentice.
Fascinating point, and one I think can definitely apply here.
Though there is a key difference – Galileo could see through his telescope the same way, every time. He also understood what the telescope did to deliver his increased knowledge.
Compare this with LLMs, which provide different answers every time, and whose internal mechanisms are poorly understood. It presents another level of uncertainty which further reduces our agency.
> Though there is a key difference – Galileo could see through his telescope the same way, every time.
Actually this is a really critical error- a core point of contention at the time was that he didn't see the same thing every time. Small variations in the lens quality, weather conditions, and user error all contributed to the discovery of what we now call "instrument noise" (not to mention natural variation in the astronomical system which we just couldn't detect with the naked eye, for example the rings of Saturn). Indeed this point was so critical that it led to the invention of least-squares curve fitting (which, ironically, is how we got to where we are today). OLS allowed us to "tame" the parts of the system that we couldn't comprehend, but it was emphatically not a given that telescopes had inter-measurement reliability when they first debuted.
LLMs can be deterministic machines, you just need to control the random seeds and run it on the same hardware to avoid numerics differences.
Gradient descent is not a total black box, although it works so well as to be unintuitive. There is ongoing "interpretability" research too, with several key results already.
Deterministic doesn't necessarily mean that can be understood by an human mind. You can think about a process entirely deterministic but so complex and with so many moving parts (and probably chaotic) that a humble human cannot comprehend.
I have been meaning to read Feyerband for a while but never did. I think Against Method sounds like a good starting point.
Did Feyerband also not argue that Galileo's claim that Copernicus's theory was proved was false given it was not the best supported hypothesis by the evidence available at the time.
I very much agree with your last paragraph. Telescopes are comprehensible.
> Did Feyerband also not argue that Galileo's claim that Copernicus's theory was proved was false
My reading of AM was that it's less about what's "true" or "false" and more about how the actual structure of the scientific argument compares to what's claimed about it. The (rough) point (as I understand it) is that Galileo's scientific "findings" were motivated by human desires for wealth and success (what we might call historically contingent or "poltical" factors) as much as they were by "following the hard evidence".
> Telescopes are comprehensible.
"Comprehensible" is a relative measure, I think. Incomprehensible things become comprehensible with time and familiarity.
> I haven't heard of that being the argument. The main perspective I'm aware of is that more powerful AI models have a compounding multiplier on productivity, and this trend seems likely to continue at least in the near future considering how much better coding models are at boosting productivity now compared to last year.
This is the new line now that LLMs are being commoditized, but in the post-Slate Star Codex AI/Tech Accelerationist era of like '20-'23 the Pascal's wager argument was very much a thing. In my experience it's kind of the "true believer" argument, whereas the ROI/productivity thing is the "I'm in it for the bag" argument.
The adoption and use of technology frequently (even typically) has a political axis, it's kind of just this weird world of consumer tech/personal computers that's nominally "apolitical" because it's instead aligned to the axis of taste/self-identity so it'll generate more economic activity.
Wow I've needed harmony for years, thanks for sharing! My dumb ass was filling out the Musicbrainz by hand for like two months before I just gave up on beets.
You'll want to bookmark Harmony even if you're using Picard since it's driven by the same database. Happily, once you add an item using Harmony, Picard can find your just-entered release almost immediately via the Release ID of the release you just created.
I find myself needing to create releases for ~10% of the albums I tag, and Harmony is a game-changer for that.
> “Worse insomnia, depression/anxiety or the use of other sleep-enhancing medicines might be linked to both melatonin use and heart risk,” Nnadi said. “Also, while the association we found raises safety concerns about the widely used supplement, our study cannot prove a direct cause-and-effect relationship. This means more research is needed to test melatonin’s safety for the heart.”
This seems pretty critical, no? I would assume that melatonin use increases with the severity of insomnia, and persistent lack of sleep has a whole host of long-term health implications. So insomnia a priori seems highly likely to cause both increased melatonin consumption and heart disease. Given how obviously confounded this causal system is, it's practically negligent to not at least control for severity of insomnia or something. Ideally they should have done a matched study or DID or something, but without any attempt to account for the confounding this study is just clickbait.
There used to be some darkweb marketplaces that were extremely reliable for psychedelics, for example "The Majestic Garden" was an old PHP message board which specialized in psychedelics and had a strict whitelist about what substances could be sold. TMG has since shut down unfortunately, but there are lots of other providers. Some "well-known" manufacturers used to hang around TMG, and if a listing on some other site mentions that it's their LSD then IMO it's a good place to start. General methodology is to get a burner laptop running some secure *nix OS (ie. Tails), acquire some Monero, navigate to an appropriate marketplace via Tor, and make your purchase. Use your real address and name for the shipping; it seems scary but in the US it's the most secure way (least likely to trigger a warrant being issued to search your package). Once that's done you head over to dancesafe.org on your normal laptop and order what they recommend for LSD reagent testing. Then you wait for everything to arrive, follow the instructions in the reagent testing kit, and if everything comes out "okay" then you're good to go.
In general it's fairly rare for someone to sell "fake" or "cut" LSD. This is for two broad reasons:
1. LSD isn't addictive in the same way as some other illegal substances, so there's much less profit in faking it. If you sell someone fake LSD and they have an awful trip they're never going to buy from you again and you've burned a profit stream. If you sell someone coke cut with meth they're probably going to come back anyways, because they're hooked on the drug.
2. The things you might adulterate or fake LSD with (ex. NBOMe) are also somewhat difficult and expensive to make, so it's not like one would really save much time or money. If you're trying to burn someone in an LSD sale, the simplest and most profitable thing to do would be to just sell empty blotter paper, rather than wasting research chemicals you could potentially sell to some psychonaut with the correct label.
With that said NBOMe and some other synthetics have been cropping up in the last 20-ish years. If someone at a party or festival is selling single tabs that's definitely the place where you might get burned with NBOMe, since by the time your trip ends the dealer will be long gone. In general it's always best to test everything you ingest, and dancesafe.org is definitely the place to get info on how to test stuff.
It's a bad idea to send darkweb packages to a PO box, as the USPS will require a lower threshold of evidence to search the package. As it is they can't inspect your mail without a warrant, which is exceedingly rare for them to procure for a random package. Having it shipped to your home, although maybe somewhat nerve-wracking, is the simplest and most secure option in the United States (for now).
After reading the blog post, it seems like there's two issues:
1. This type of question (return a desired emoji) requires a high-degree of "accuracy" on a single token. Contrast that with more typical LLM tasks which tend to emphasize more holistic "correctness" of multiple output tokens.
2. The (mode of the) token probability distribution converges to a "hole" in the token corpus, but the model is designed to "snap to" the token nearest the hole. So it returns the wrong emoji. Normally this isn't a problem, since token embeddings are constructed so that things near the "hole" have similar semantic meanings, so perform equivalently in most sentences. But this is where Issue 1 rears its head: exact 1-token accuracy is the performance metric for evaluation, so something "similar" to a seahorse emoji is as bad as something totally unrelated.
These two core issues are particularly problematic as production models are fine-tuned to be "self-reflective", so the model reasoning chain then causes it to keep retrying the task, even though the problem is ultimately an issue with the tokenizer/token embeddings. Some models are capable of converging to the "correct" answer which is to spit out a sequence of tokens which can be read as "none exists"; this is probably heavily influenced by the prompt ("is there a seahorse emoji" vs. "show me the seahorse emoji").
I think the real way we need to reason about this is via the topology(/homology) of the underlying embedding space; seems that our current tools assume a Cauchy-complete token space. In reality some tokens simply are undefined. While intuitively that seems rare for natural spoken/written language (as an undefined token is a semantic meaning without a word, and people tend to just make up new words when they need them), in the world of "hard languages" (coding, math, pictograms/emojis) these topological holes are actually meaningful! A coding language might have a truly undefined token, even though it is semantically similar to other tokens in the corpus. Moreover the topology near these holes can be super misleading (everything is infinitely continuous up until you fall into it), so it's basically the worst corner-case for the kinds of iterative gradient descent algorithms we use to build NNs. It seems like we need a richer set of constructs for representing language tokens than Banach spaces; a super thought provoking area of work for sure!
Wow this is extremely cool/impressive, but if my manager asked me to implement this I'd quit lol. The "state" headaches alone seem like a nightmare, nevermind all the whacky linear algebra you're going to have hand-roll (Like does Postgres even have a matrix type?? Did you have to implement matrix inversion in SQL from scratch?? I get nauseous just thinking about it.)
edit: I guess in 2D a lot of this becomes simpler than in general high-dimensions.
A while ago I read "Against Method" by Paul Feyerabend and there's a section that really stuck with me, where he talks about the "myth" of Galileo. His point is that Galileo serves as sort of the mythological prototype of a scientist, and that by picking at the loose ends of the myth one can identify some contradictory elements of the popular conception of "scientific method". One of his main points of contention is Galileo's faith in the telescope, his novel implementation of bleeding edge optics technology. Feyerebend argues that Galileo invented the telescope as primarily a military invention, it revolutionized the capabilities of artillery guns (and specifically naval artillary). Having secured his finances with some wealthy patrons, he then began to hunt for nobler uses of his new tool, and landed on astronomy.
Feyerabend's point (and what I'm slowly working up to) is that applying this new (and untested) military tool to what was a very ancient and venerable domain of inquiry was actually kind of scandalous. Up until that point all human knowledge of astronomy had been generated by direct observation of the phenomenon; by introducing this new tool between the human and the stars Galileo was creating a layer of separation which had never been there before, and this was the source of much of the contemporary controversy that led to his original censure. It was one thing to base your cosmology on what could be detected by the human eye, but it seemed very "wrong" (especially to the church) to insert an unfeeling lump of metal and glass into what had before been a very "pure" interaction, which was totally comprehensible to the typical educated human.
I feel like this article is expressing a very similar fear, and I furthermore think that it's kind of "missing the point" in the same way. Human comprehension is frequently augmented by technologly; no human can truly "understand" a gravitational wave experientially. At best we understand the n-th order 'signs' that the phenomenon imprints on the tools we construct. I'd argue that LLMs play a similar role in their application in math, for example. It's about widening our sensor array, more than it is delegating the knowledge work to a robot apprentice.
reply