delis-thumbs-7e

Nice one. I need to read this later.

if history and philosophy are so important

Jesus Mary and Joseph we are cooked.

Through Apollinaire, Picasso contacted the poet’s ex-secretary, Honore-Joseph Géry Pieret, who was ready to steal artifacts for a reasonable price. In 1907, Pieret broke into the Louvre and took several sculptures with him. Months later, Picasso would reveal his ground-breaking Cubist work Les Demoiselles d’Avignon which was heavily inspired by Iberian and African sculpture.

https://www.thecollector.com/famous-artists-turned-crime/

https://en.wikipedia.org/wiki/Whataboutism

Thank you. I need to try if this helps my mom survive smart phones.

The bloody kid can take up some nice linen canvas, few sticks or charcoal, oil paint, turpentine and bloody paint his selfie! If it was good enough for Caravaggio, it is good enough for you!

How many languages have ypu learned with your immersion technique? Maybe you should search a bit of research on how successfull your immersion is for adults learning a new langauge. Perhaps then take a look what research says people who actually learn new languages (not youtube polyglots, people who have to do it for work etc.) do.

You are completely right, and hours spent (in class) is how say US Foreign Service measures how quickly you get somewhat competent in a given language. There’s a lot of hoopla about just immersing yourself to the language and so on, but for most people achieving B1 means going to classes several hours every week for quite awhile.

Looking at the content, if you are familiar at all with Python and basic programming, this provides very little new. I sometimes have to stuff massive Excel-abominations with 50k+ rows and rip data I need out of them with Pandas, but it only requires reading Pandas documentation (which is very good) a bit.

But perhaps this might be good if you know no programming and want to make your life easier.

Why can’t you grow 10cm of height each year? Toddlers do it very easily, and they don’t even go to the gym. Maybe you just don’t work out enough? Average 40 year-old man should be about 4 metres tall, otherwise he just gave up.

It just requires turning your rational mind off, immersing yourself in the play/sport and trusting things will work out, something which more rational/analytic people tend to struggle with.

I do typos and grammar mistakes in four languages (only two of them fluently) and it’s pretty easy to learn basic vocabulary, read a bit of a comic book and order coffee with a thick accent. A lot of “polyglots” who can do basic phrases in 11 languages and make videos about it on YouTube. There’s people who genuinely speak plenty of languages fluently, but they are quite humble about it. Reaching B1 after completing A2 is 200-300h of study, preferably in a class. That’s a lot for a working adult with a family.

However, there is plenty of people who think that they speak a language because they can order a coffee in some language close to theirs and have never lived in a foreign country for any extended period of time, ie. don’t know what they are talking about, and arguing from the position of complete ignorance say “just learn the language”.

The idea that there’s one country with one language is very recent actually, part of the nationalistic idealism of 19th Century. Most of the history, as it is still today, any nation is a mishmash of cultures, languages, dialects of various kinds and origins, there of course always being some official one (used to be Latin, then French, now it’s TikTok-memes). The fact that pur systems cannot facilitate that multitude is a shame and I believe those systems that can will succeed.

None of this people opining “just learn the language” have learned a second language while working as adults, let alone learned German. You can get to A2 level pretty easy (in most indoeuropean languages at least), but jumping to B1 can seriously be a year or more of studying. You have to be able to handle basic daily situations in the given language and understand what is said in a TV or Radio show. With practice you can get there especially if you live in the country and force yourself to speak the language, but easy it is not.

I disagree on your definition of deep structure. We can indeed deep structure, that is my whole point, without it Finnegan’s Wake would be impossible to read. This really is something LLM’s cannot do, at least for now. And if you look at the paper I linked, if they had a world model they would not get confused on mathematics when we change the values of variables. So there seems to be plenty of things LLM’s cannot do.

And basically the burden of proof is on your side. As often is the case, discussions on AI start to resemble discussions centred on religion and belief in god. Your only line left here is “yes but you cannot prove LLM’s are conscious” - which I can’t, since proving that Unicorns do not exist is pretty senseless activity.

So basically I’d say LLM’s are stochastic parrots (or whatever term you want to use) until we can very definitely prove that they indeed think and have some sort of consciousness.

All physical exertion causes pain and discomfort, otherwise it isn't exertion.

This is just ridiculous. If this is actually true to ypu, please seek medical care immediately. You might not have long left.

It addresses the point exactly. I doscuss my experience to show I quote know what I am talking about. Any amount of exercise is pnly good for long run if you do it consistently, with enough intensity and for long time. Your 5min ab workout that you manage to keep up for two weeks is basically meaningless. 5000 steps, 15min workouts, 30min jogs, there has been plenty of this fads and no, they do not get people moving in the long term. Forcing yourself to put on running shoes every morning works up to a point when it doesn’t - you get injured, ill, meet someone, Christmas, summer vacation abroad, new job, anything. And you won’t go back, because you hated running in the first place.

What does, is getting a hobby where you move. It might be chasing butterflies or Freestyle Wrestling, doesn’t matter as long as you think it is fun.

I don’t know why you argue when I’m basically giving a cheat code how to start enjoying sports and exercising. If you enjoy something you do it more, what is here to argue?

Yes. I ran a half-marathon few years back when I used to box actively (didn’t compete, but sparred competitors). I woke up early 3 times a week to do 5km run no matter the weather. These days I just lift heavy weights. I think I know every motivational technique there is at this point.

What matters is that you move, consistently, every week, for years, for rest of your life. I’m kinda person who wants to just lay on bed and eat potato chips drinking beer. I used to start some regime or another, always getting bored after few month. Then I just found types of sports and exercises that worked for me and company they are fun to do with, and guess what, I need almost zero motivation to lift some several thousands of kg’s of iron every week.

Consistency, intensity, motivation. Best motivation is fun, especially fun with other people.

Obviously there is disabilities that make almost all physical activity impossible. I don’t see your point, I should have made a disclaimer ”except of course people who have amyotrophic lateral sclerosis, [list every possible disabling condition]?”

But plenty of people with very disabling conditions became those world class athletes, or live life where they use their bodies to the extent they can. I bet it is because they have found a way to move they truly love. If you do something you like doing, you do it more.

I suppose that if your muscles grow rapidly, you could call that inflation of some kind? Connection to cancer still unclear though.

There’s even Paralympics in this world. People with disabilities do sports, although obviously it complicates things.

How about you instead of trying to “get away with” as little exercise as possible you found a way to move your body that you genuinely enjoy? One should not optimise away sleep, nutrition and exercise. Yes we live in a ridiculous fitness and gym boom, but weight training might not actually be a good fit to most people. I love having to psych myself to lift big amounts of steel, I’m ok with bit of pain etc. - mostly people are not. There’s thousands of sports and ways to move out there. You don’t like running, well don’t. Go dancing, hiking, play a team sport, do tai-chi, yoga, goddammit change a hobby every two months, it doesn’t matter. Just do it and do it consistently every week.

And what best helps you in that is enjoying what you do, as well as doing it with a friend.

I am no linguist, but I believe this is referred to as surface structure and deep structure. What you describe is a line of text that is grammatically somewhat adequate to pass as readable and you treat all text the same. when we read the text, we decipher meaning out of according to word references and syntax - just like with Python or C++. However, if this were solely the case, we could not read Finnegan’s Wake. Probably vast majority of modern poetry would be unread by anyone, as would be pretty much all major works of philosophy Kant onwards. Deep structure is what according to Chomsky et al. gives meaning to the language, ie. somewhat logical structure behind the mere words. English word strict has order, need it but actually not does. You skibidi rizz swag grok brah also, barely. We use language in the extended meaning of the word to create a model of the world and somehow the past riverrun skibidi transmits that model to others. This is what I believe Bender also tries to say in their paper.

Now, you could claim that LLM’s have this deep structure, create models of the world and are basically just like us, and certainly many here are adament that this is the case, being aghast how someone can “insult” LLM’s by calling them parrots. However, there really is not much proof to back up this belief. Usually LLM’s seem to copy existing surface structure from whatever source, and when it deviates from these patterns, it usually becomes incomprehensible. There’s much hoopla about LLM’s solving hard maths, but it seems even there they are mostly generalising from vast amounts of training data, rather than actually reasoning: https://arxiv.org/pdf/2410.05229

You know monkeys can swim? Do you know what happens to a monkey 2-6km under the surface? Monkeys can also climb trees, very high. Monkey still no condor, sadly.

If you want to see words form a shape I could point you towards concrete poetry, but I guess there is no point. Joyce wrote Finnegan’s Wake for 17 years and although superficially it seems complete gibberish, trodding through it you find meaning to words that are in no dictionary, sentence structures alien to English, etc. but still you are able to understand it, and perhaps some way the mind that produced it. So I disagree with you, we can observe each other’s inner life. It is always unexpected, strange, exciting, but always rooted to our shared experience or what it like being a very big and confused ape.

LLM’s are usually unexpected only when they malfunction and sprout same letter again and again etc - hardly a literary masterpiece. They make very easily recognisable patterns that we can use as helpful tools, but in the end they are devoid of any meaning apart from what we give them. Of course one could say same about art and all language, but I think there still is the fact that we apes somehow recognise each other. And besides, we do know the internal functions that drive the parroting. It is admittedly bit tricky, but in no way as magical as people purport it to be.

I think this is completely misleading. If your employer asks to redact your paper because it ignores relevant research, naturally you want to know what research they are talking about and on what grounds, and also which researches reviewed the paper (probably assuming there was none, review process was used as an excuse to silence criticism, G has done it many times).

(BTW, quite bold to say input data from Reddit and 4Chan is how “normal” people speak. There is a lot of language in the training data of any model you really do not wish your application to use ever.)

There’s a lot of confusion here on what parrots exactly are capable of doing and assumptions how smart they are. I also don’t quite see what you mean by behaviour in respect to LLM’s, possibly “agentic” tool usage? Because I’m fairly certain you can explain their behaviour as “stochastic parroting” combined with very strong program, ie. harness, to interact with other systems. And perhaps this all is “some kind of intelligence”, but then you have to be very careful with what you mean by intelligence. It becomes a terribly slippery surface described like this and you see, who knows where this gradient descent takes us if we are not careful.

So personally I would still rather undersell it. But parrot or not, it is still a terribly useful little bird indeed.

Yes, I am: https://en.wikipedia.org/wiki/Alex_(parrot)

A large, monied cross-section of the world

I asked Mistral, and it guestimated that Altman, Thiel, Musk, and Hoffman had like $20.3B together when they founded it. Sound to me that the founding of OpenAI was exactly the point when the monied world threw its dollars behind AI.

I really appreciate the effort you put in to this post. Posts like these are what makes HN great. Thank you.

Why is it not indicative of what LLM’s do?

Well in a sense that monkeys and Great Condor inhabit the exact same environment.