I love this kind of site. Not that you can't scour up the content yourself using Google, but here's the thing... when you're trying to "teach yourself" something complex, one of the challenges you run into (in my experience) is understanding the order of the dependencies for prerequisites and sub-topics. And nothing is more frustrating than trying to read a book or a paper and feeling totally lost because you're missing something that you should have learned first, but you don't even know exactly what the "something" is. :-(
As someone who considers himself something of an autodidact and who is always trying to learn new stuff, I really appreciate a resource that can point at a bunch of stuff and say (even in an approximate sense) "do this in this order". It really does make a difference.
Bingo. This is in complete contrast to visiting wikipedia where, for example, if I go to any mathematical concept, I seem to tumble down a series of holes that usually falls out somewhere much more abstract that what I need (usually topology). This is not to say topology is not important, but clearly becoming an expert topologist before learning wavelets (for example) would not be the most efficient way to learn the latter.
To borrow Prof. 't Hooft's imagery, I imagine that GAI would be right at the top of a skyscraper whose lower levels consist of mathematics, cognitive science, and computer science at increasing levels of sophistication.
Beginning your study of cognitive/computer science with GAI could be frustrating and fruitless, just as would be beginning your study of physics with superstring theory.
There are probably well-written and interesting GAI books/articles for laypeople, though, analogous to Brian Greene's excellent books on string theory.
Your response is totally understandable given my (very clumsy) comment. Let me make myself more clear; I'm in the process of reading up on what material one should study to set oneself up to ultimately be able to consider AGI research as a career choice.
You say "mathematics, cognitive science, and computer science at increasing levels of sophistication.". That's probably true, I'd recon a fair bit of biology, chemistry and physics is most likely also needed.
My question, though, is what specific branches of said subjects?
Don't know of one off hand. I know a few curated lists related to specific AI/ML topics have shown up here on HN (mostly NN and DL stuff,IIRC), but I don't remember one that covers AGI. There's definitely a TON of great AI content on the 'net, but yeah, a nice curated guide to help work through the important stuff, in order, would be very helpful.
The most profound thing that the internet can do for the world is create a single monolithic repository of documents that is designed to take a dedicated reader from basic principles of math to a high degree of understanding in any subject. This is such an easy task compared to all of the other projects that happen on the internet, it is astonishing that we don't have anything like it yet. Wikipedia comes close. Khan Academy might become what I just described after they add more content. But nothing really hits the nail on the head yet. If Facebook spent a few billion on it then we would have it. It's so fucking simple. I would trade in Instagram many, many times over in exchange for such a repository. It is one of the easiest things to do and out of all the things we could do has the greatest impact on society and yet it has not been done. Amazing.
The most important part of this is to do the problems!
After a PhD in theoretical physics then 20 years as a s/w developer I decided to get some of my chops back. You can read and nod about stuff but until you face a blank sheet of paper you don't know if you really understood.
Before I clicked that link I thought the parent might be complaining for no reason. Then, and this was probably just a coincidence, I saw the page and had a pressure buildup in my right eye that felt like a headache.
I'm happy to see that there is someone who is trying to help those who don't want to or can't go the traditional route of undergrad => grad school. Certainly I'm sure there are many others like me who have a strong, but amateur interest in a subject. Here I mean amateur in the sense that it is unpaid and done only out of love, as opposed to some measure of skill.
Sadly I have no intention of studying theoretical physics. If anyone could point me towards something for theoretical computer science and algorithms I would be most grateful
Hey thanks for the suggestion. I actually use this strategy already, but the issue is that this only tells you what the general stuff most people will learn is. As far as I know this doesn't help with finding what the current problems are in the field.
However, I think that professors generally are very willing to help people who are interested, even if you aren't a student. Actually as far as I can tell usually they don't even ask if you are a student.
In addition to that you also can typically know how an undergraduate would progress in the courses from the universities as well, through stuff like graduation plans etc.
Aside: that phrase "someone who is trying to help" makes me wonder whether you know that _this_ someone is a Nobel laureate (shared with Veltman in 1999), as he casually mentions in the text.
I did read it so yes I did know. I just feel like most active researchers would be able to point people in the right direction, so the fact that he's a Nobel laureate wasn't particularly relevant.
I speak reasonably fluent Greek (family language), and took a course in Classical Greek when I was in school. My professor explained that the pronunciations used in mathematics (tau, beta, mu...) differ from the Modern Greek versions (taf, veeta, mee...) because the former are the (conjectured) pronunciations used in the classical language. A large source of evidence for the pronunciation of Classical Greek is the transliteration of Greek words in to other languages during that period, especially into Latin. The whole language is pronounced very differently, not just the alphabet.
As a side note, that class was absurdly easy because I already knew almost all the vocabulary. The classical grammar is quite different though, and more complicated.
'Tau' is the ancient Greek way. My classical languages teacher was able to converse in ancient Greek and Latin, but couldn't make head or tails of modern Greek when he visited.
Has anyone here ever read or worked through Hrabovsky and Susskind's _The Theoretical Minimum: What You Need to Know to Start Doing Physics_? Same premise as this, I gather, but a whole book.
For anyone who wants to read this. The book was rebranded. It's "Classical Mechanics - TTM" (this used to be called TTM only) now and there's also a "Quantum Mechanics - TTM". I have both of them on my nightstand but only browsed so far. Seems good at first glance. Susskind's Stanford classes for laypeople were great so I expect the books to be good as well :)
The TTM books grew out of the Stanford Continuing Studies courses by Susskind.
These were recorded by Stanford and uploaded to Youtube: http://theoreticalminimum.com/courses
There is very much more content in these lectures than is covered in the two books.
Unfortunately, the section on Plasma Physics is rather sparse.
Introduction to Plasma Physics by Francis F. Chen and Fundamentals of Plasma Physics by Paul Bellan are good next-step texts, and this crowd in particular might enjoy Birdsall and Langdon's Plasma Physics Via Computer Simulation
Also on the computational topic, Kane S. Yee's 1966 paper "Numerical Solution..."
Wow, I originally clicked on the article thinking it was tongue-in-cheek, but it's fantastic! Glad to see that someone is taking time to help others (like me)
Regarding his comment about hearing from amateurs who think they've solved the world, that is something I've heard from a lot of physics professors, that they get a lot of emails with crazy stuff like this:
http://i.imgur.com/QTt6ZTq.gif
Why does that happen in physics and not chemistry, bio, econ, comsci or any other subject?
It happens in math and theoretical CS all the time. Proofs of P != NP (or the opposite!) are particularly popular.
In applied CS, instead of emailing professors people just post their stuff on the internet. Sometimes it gets wildly popular. How do you think PHP came around?
Everyone's an armchair economist so I wouldn't count econ out.
I think that math/physics are more accessible to the layman. I know where to go if I want to get deeper into this subject.
I really don't know how to start becoming an amateur chemist or biologist. I've wanted to learn more about synthetic biology and gene therapy, but it's not as simple as cracking open a math book or going to a dev bootcamp. I wish these resources existed though! My best guess is because exploring chem and bio have bigger resource requirements like lab space.
I think math/physics are less accessible to the layperson, but more subject to Dunning-Kruger. There are so many popular accounts of quantum this and string that in the media that a lot of the population seems to feel qualified to hold an opinion on anything quantum-y. ("It's on TV" => "I'm encouraged to have an opinion about it.")
In reality it's literally impossible for most people to imagine how hard the math is - because an undergrad physics or engineering degree is barely even a warm-up for it.
I'd love to see a TV show that made it clear just how challenging the math is without reducing it to the usual storytelling.
Popularisation can be great, but maybe it would be good to get a realistic appreciation for the raw version into the public narrative.
Biology, and I Assume chemistry as well, are in practice all about lore; the peculiarities of particular entities are where the action is & theory is a relatively small thing that falls out of that. So not only do you need a lab to do experiments, you need to be hanging around a lab to pick up all this disjointed lore.
Oh it happens all the time in econ - although one could argue that economics at the moment does deserve some of it. You regularly get emails and even the occasional book, linking to yet another youtube misconception of how money, banking and the economy work. I usually reply pointing out the time at which the first provable mistake is made, and I've yet to get over 5 minutes. They do get extra points for being vaguely anti-semitic before that point.
As a former professor it does happen in biology. Whenever some crazy person would ring the switchboard in the biology area they would pass it on to me - I am not sure why, maybe it was because I was relatively patient with the people who called. Lots of interesting conversation, but little science.
The author of the submitted website is a native speaker of Dutch, but his English looks fine to me (a native speaker of English). I think his choice of emphasis, which I didn't try to reproduce when I submitted the website here on Hacker News, makes the point that people can become mere amateurs or even cranks[1] about theoretical physics unless they prepare carefully. I have seen a few examples of theoretical physics cranks[2] over the years.
[1] I mean "crank" in the second sense shown at the Merriam-Webster dictionary site.
I agree that use of "good" is questionable here. One would think that without the skills mentioned, one wouldn't be a theoretical physicist at all. (Is every person who lacks these skills a "bad theoretical physicist"? Only in a vacuous sense.)
That said, I think we should interpret "how to become a good theoretical physicist" as "how to gain the competencies that are expected of s theoretical physicist."
That's is what he means by good: one with enough competence in the field to do actually useful work. If you manage to do useful work in physics, that work could translate into a Nobel Prize just as easily as any other useful work.
A bad theoretical physicist is one who only understands enough to sound smart to their friends. AKA, a crank.
I am not sure that you will become a 'good' theoretical physicist after reading those. (I have knowledge of most of the items he mentions) You will have a good overview of many mathematical tools that are used in physics but that's about it. Personally I prefer to read the stuff the 'bad' theoretical physicists write. Read Penrose or Bohm for what 'good' theoretical physicists think of existing orthodoxies and their many failings.
"set up only for those who wish to become theoretical physicists, not just ordinary ones"
--- use of 'ordinary' instead of 'experimental' is intentionally condescending, hopefully for sarcastic effect
"very best, those who are fully determined to earn their own Nobel Prize. If you are more modest than that"
--- I'm not sure if this is actual encouragement or a joke. But the following:
"finish those lousy schools first and follow the regular routes provided by educators and specialized -gogues who are so damn carefully chewing all those tiny portions before feeding them to you."
--- Must be a joke. Nobel prizes are not awarded to individuals these days, they go to collaborations. There is no way to get into a collaboration, or likely a PhD program without an undergraduate degree in Physics. Legally not allowed in many European countries to join a PhD program...
"More than rudimentary intelligence is assumed to be present, because ordinary students can master this material only when assisted by patient teachers."
--- if this isn't sarcasm, then this guy is a monstrous prick who just decided to insult the 'rudimentary' intelligence of all Physics undergrads who deigned to actually attend college. Again with the 'ordinary'.
The opening paragraph continues the authors apparent disdain for school in general:
"But what if you are still young, at School, and before being admitted at a University, you have to endure the childish anecdotes that they call science there? What if you are older, and you are not at all looking forward to join those noisy crowds of young students?"
Which is a bit humerus considering the importance he himself places on fundamentals with:
"solid foundations in elementary mathematics and notions of classical (pre-20th century) physics. Don’t think that pre-20th century physics is “irrelevant”"
If it isn't sarcasm, then I have no idea why the author has decided to be so acerbic in his notation. I'm giving him the benefit of the doubt that he doesn't actually mean to insult basically the 99% of physicists who don't show up in "Brilliant Mind"-esque movies.
edit:
From his CV:
* Gymnasium-Beta, Dalton Lyceum, The Hague, 1964
* Physics and Mathematics, University of Utrecht
* Kandidaatsexamen N1, 4 July 1966
* Doctoraal examen Theoretical Physics, 10 October 1969
* Promotie (PhD thesis) on the subject "Renormalization Procedure for Yang-Mills fields", 1 March 1972
i.e. He did basically all this stuff he is brushing off as worthless. As in going to classes, with teachers, at an undergraduate university, before applying for a PhD program.
Well, that is one way to read it, I suppose. Another is to take the second paragraph as the theme of the piece: if you don't have a grounding in all of this stuff, your reckonings and supposings are probably not as brilliant and revolutionary as you think they are.
The author isn't expressing his disdain for school, but the disdain he perceives in the people who keep e-mailing him with the simple answers to deep questions. "Okay," he says (I'm paraphrasing wildly here), "you may be too cool for school, but if you don't have at least a firm understanding of these basics, you won't even be able to understand why you're wrong when you, quite inevitably, veer off into the muddy mire of misunderstanding."
He is most assuredly not denigrating experimental physics, nor is he saying that the ordinary educational process is a waste of time and effort. (This is not meant to encourage autodidacts in any way.) He's even hinted that the "lies to children" -- those models that seemed to work for a very long time in the pre-GR/pre-quantum world -- aren't nearly as useless as people might think they are, nor are the scientific processes that led to those models.
So, yes, he is being quite sarcastic here, but the sarcasm is pointed in a completely different direction.
You are misreading anger where there is slightly ascerbic honesty:
> "set up only for those who wish to become theoretical physicists, not just ordinary ones" -> -- use of 'ordinary' instead of 'experimental' is intentionally condescending, hopefully for sarcastic effect
If you'd finished the sentence, you'd see: "but the very best, those who are fully determined to earn their own Nobel Prize. "
He is constrasting "ordinary" with "the very best". He isn't talking about experimental physics -- that's not his field.
> "finish those lousy schools first and follow the regular routes provided by educators and specialized -gogues who are so damn carefully chewing all those tiny portions before feeding them to you." --- Must be a joke. Nobel prizes are not awarded to individuals these days, they go to collaborations. There is no way to get into a collaboration, or likely a PhD program without an undergraduate degree in Physics. Legally not allowed in many European countries to join a PhD program...
The beauty of theoretical physics (and math) is that you don't need any expensive equipment or overpriced tuition. Just books and people to talk to. Both of these are available free (except some books cost $$$), on the Internet. That's enough to get started.
> "More than rudimentary intelligence is assumed to be present, because ordinary students can master this material only when assisted by patient teachers." --- if this isn't sarcasm, then this guy is a monstrous prick who just decided to insult the 'rudimentary' intelligence of all Physics undergrads who deigned to actually attend college. Again with the 'ordinary'.
It's a simple fact: Smarter folks can get farther without as much guidance from teachers.
> The opening paragraph continues the authors apparent disdain for school in general:
"But what if you are still young, at School, and before being admitted at a University, you have to endure the childish anecdotes that they call science there? What if you are older, and you are not at all looking forward to join those noisy crowds of young students?"
Common K-12 schools in many countries generally teach science and math very poorly. This is not a controversial statement.
Colleges are full of 18-22-yr-olds, many of whom are there to party; not a comfortable environment for some 30+ year olds.
> Which is a bit humerus
You seem to have a bone to pick.
> considering the importance he himself places on fundamentals with:
"solid foundations in elementary mathematics and notions of classical (pre-20th century) physics. Don’t think that pre-20th century physics is “irrelevant”"
What does that have to do with the high school and college learning experience?
>i .e. He did basically all this stuff he is brushing off as worthless. As in going to classes, with teachers, at an undergraduate university, before applying for a PhD program.
The into goes to lengths to explain the sort of person who doesn't fit into school, and would benefit from an autodidactic alternative. For everyone else...there's school. I don't know why you have such a chip against the possibility of thriving outside of college... too much student loans got you sour grapse?
I'm an experimental physicist who did an undergraduate before my PhD and had to fight tooth and nail for my grades and did all that petty nonsense like going to class and after school tutoring and such. I'm not a rockstar, but I'm not a joke either -- I would give myself a "B" rating -- I definitely don't struggle for fellowships, but I'm not Nobel material.
I know from the inside that _these_ actions are the most probable path to success. If some hot shot high school student reads his note, I can see them deciding that they are "brilliant mind" material and then start pursuing this course of action seriously. In my opinion that is a borderline surefire path to failure. College campus is _the_ place to learn physics: you learn physics from casual bar discussions about the fluid dynamics of beer, telling a pretty girl why stars twinkle and planets don't, explaining to stoners why cigarette smoke is blue off the cig and then grey when you exhale, pre-exam pizza binges, adding strobe settings to your dorm lights, and late night campus laser graffiti with your mates and professors. Not from lone wolfing down a textbook with a blackboard.
It's perfectly possible for him to convey his message without "ordinary", "childish", "noisy", "rudimentary", "tiny portions" etc. And if he's just putting together a guidebook for entertaining your interest in Physics, then that's cool. But saying this is the way to get a Nobel is ridiculous... hence my original comment about dripping sarcasm.
Also my original comment was just getting down voted (and someone asked for clarification) so I expounded.
You're right that Professor 't Hooft is being sarcastic but you've got the wrong target.
He's not aiming this at practicing physicists. He says early on that, if you want to be a physicist, you should enroll at a University. He says in the Questions section that you'll eventually need a degree.
This is aimed at people who can't or won't. The sarcasm is intended to appeal to the "well intended but totally useless" amateur physicists that regularly send him crackpot theories. It's a common refrain from such individuals that they are held back by the establishment.
What he's trying to convey is the amount of information that one needs to master to be taken seriously as a theoretical physicist. His intention is both to put off those that have read a popular science book and now know it all as well as encouraging those that are actually serious by giving them some structure to their learning.
Well I want to be a good physicist and I appreciate that it will take me years, maybe decades of reading. That's what's necessary to catch up to what we know. It's tough going but equally rewarding to learn.
I can tell you of my own experiences. I had the extreme luck of having excellent teachers around me. ... It helped me all the way to earn a Nobel Prize.
But I didn’t have internet. I am going to try to be your teacher.
It is presently set up only for those who wish to become theoretical physicists, not just ordinary ones,
but the very best, those who are fully determined to earn their own Nobel Prize.
This is a site for ambitious people.
I am sure that anyone can do this, if one is gifted with a certain amount of intelligence, interest and determination.
As someone who considers himself something of an autodidact and who is always trying to learn new stuff, I really appreciate a resource that can point at a bunch of stuff and say (even in an approximate sense) "do this in this order". It really does make a difference.