Wolfram might be a genius, but I personally find his writing and lectures extremely dull. The only thing he seems to write is how everything is "computational", and how we need to embrace his idea that computing is what the whole universe is about. For example in this blog post alone, we can find this following word count:
Throwing all these words around may sound smart but they lose any meaning or relevance that they were supposed to deliver if being overused in such a larger-than-life manner.
Welcome to Stephen Wolfram. Wolfram is madly in love with his own ideas, even when those ideas are not nearly as original or world-shakingly profound as he seems to think they are. That he is a genius does not mean he is not in a state of arrested development; prodigious genius like his can impede one's development because people around the genius are so in awe of them that they are unwilling to give them the much-needed corrections and reality checks it takes to grow.
It's very hard to teach someone whose cup is full. Very smart people can rationalize their bad ideas faster than other people can talk them out of them. But that doesn't mean you've actually explained yourself. If you can't explain things to other people, do you really understand it yourself?
At this point Wolfram is lost to us. "What the hell are you talking about, Steve?". I was just fantasizing about resurrecting Richard Feynman and having him ask Wolfram this question but it turns out I don't have to:
In a letter from Richard Feynman to Stephen Wolfram:
> You don’t understand "ordinary people." To you they are "stupid fools" - so you will not tolerate them or treat their foibles with tolerance or patience - but will drive yourself wild (or they will drive you wild) trying to deal with them in an effective way.
> Find a way to do your research with as little contact with non-technical people as possible, with one exception, fall madly in love! That is my advice, my friend.
> If you can't explain things to other people, do you really understand it yourself?
I hear this said every now and then and I don’t think I agree with it.
A lot of knowledge builds on other knowledge. And the other knowledge might be required in order to understand the knowledge that one wishes to share.
So then either you have to explain it all, or you will have to simplify dramatically.
Few people will want the full explanation — it would take much more time to explain than they would ever want to spend listening. And I am not saying that to criticize those people. The same goes for myself — there are very many things I wish I could understand deeply. But there simply is not enough time. Specialization is necessary. We must pick the few things that interest or have the most use to us, if we wish to have a chance to truly understand any topic deeply at all. And even in the topics we choose there will be a lot of subtopics that we won’t ever have time to understand as fully as we would desire.
Even if they did want to listen the amount of information could be so huge that it might take several years to explain all of it. And so it can’t be done.
So instead we have to simplify. A lot. Unfortunately, when we simplify we often end up skipping crucial information and might even communicate the knowledge wrongly. In the cases where at least we don’t communicate it wrongly there is still a significant risk I think that one or several of
- the truth,
- the value, and/or
- the utility
of the knowledge we try to share is not possible to see without a lot of context that would take too much time to explain.
I have seen time and time again other people jump from
- the fact that someone was unable to explain something in a satisfying way
to
- concluding that the person that tried to explain it doesn’t himself/herself understand it
When really I think the real conclusion should instead be that it is impossible to determine at that present time whether what is being said is correct. But that neither means that the other person doesn’t know, nor that they do know!
I think the fundamental problem is that we keep insisting that things should be simple when the reality of the matter is that they just aren’t. Or perhaps there is some simpler answer but the only path known by the person to understanding the knowledge they possess is based on the knowledge they learned before they learned the knowledge in question. Then it could be possible for that person to boil it down to something that does indeed both hold truth and is simple to explain. And perhaps that it is what it means to understand something at the deepest possible level — to have the explanation worked out so that it builds on the fewest and most simple prior knowledge required. But to sit down and make those simplifications could also take years. And unless it is the job or the desire of the person holding the knowledge to spend all of that time they could spend on other things just to work out the simplest possible explanation of the knowledge then it won’t make sense for them to do so.
There are a few things more I would like to say about my thoughts on this too but I think this comment of mine has probably turned into a big wall of text already. So instead I will conclude with saying that I think that the ability of one person to explain something to someone else
- relies on their own understanding yes
but
- it relies to a much greater extent on the amount of knowledge that both parties already share
and
- a lot of things in daily life builds on things we all know already, and does not extend it too much, so it is quick to explain and simple to understand
but
- some knowledge is so far removed from shared knowledge, at least by the path to it known by the person trying to explain it, that it is for all practical purposes impossible for that person to explain to the other
You highlight a common fallacy that is repeated across the internet: "If you can't explain to me why I'm wrong, then I must be right," or something similar. But as you point out, this assumes that an explanation can be written in just a few paragraphs, which usually isn't the case. It takes years of study and experience to become an expert in a given field, and as a result, it becomes too difficult to accurately convey the necessary information to get a point across.
That reminds me of the movie "Bohemian Rhapsody" where Freddy Mercury splits off from Queen to write music independently.
But he returns saying:
"I went to Munich. I hired a bunch of guys. I told them exactly what I wanted them to do... and the problem was... they did it. No pushback from Roger. None of your rewrites. None of his funny looks. I need you. And you need me."
You might have seen his table analogy for abstractions, where the concept of a table was an abstraction defined by a set of nouns and adjectives. Yes, "make everything computational" because we all have not yet come to understand the abstract world that he is foretelling as well as we understand the word table.
For another analogy, just because you frame a house with one material e.g. wood, doesnt mean that its structurally unsound.
I do hope somebody who can write is reading his stuff carefully or at least insightfully. He seems like a great character to inhabit a science-fiction story. It would be a pity not to see him there.
Is `hashbrown::HashMap` now completely ported/identical with `std::collections::HashMap`? I just compared their performance (on yesterday's beta channel) on a hashmap with over 200k keys and didn't notice any relevant speedup.
Atlas is built to form a registered Delaware Corp with an agent and a corporate bank account in the US. If your goal is to be acquired or go the traditional startup route then it's perfect. If you just want to roll a new entity for legal/financial protection it is akin to combating mosquitos with a grenade.
If you just want some legal/financial protection, most countries have a simplified way to form a basic company (in the US it's LLC) that you can easily file for yourself, and usually the taxes are easier then say a C corp.
Atlas was designed to correctly setup a company to become acquired or go public. You simply don't need those features and the complexity included on the cap table, taxes, etc.
> Apart from the fact that nobody wants to risk having false positives, there’s a simple argument as old as machine learning itself: whatever a human can do, a machine can be taught to do. Humans have no problem correctly interpreting adversarial examples, so there must be a way to do it automatically.
So by all means, the authors (and a vast majority of researchers) seem to be confident that ML/DL is the road to AGI, hence can "solve" human intelligence (given it is computational)? For how long are we gonna drag the adage that mimicking a human (Turing test) is equal to reaching human levels of intelligence?
It is not true that whatever a human can do, a machine can be taught to do. The human must have insight into HOW they do it in order to teach it, or otherwise come up with some new algorithm. There are a large class of things humans do that they don't understand the mechanics behind, and for which there also aren't algorithms.
I'm not talking empathy or philosophy. How about just folding laundry. Not just one type, not in a controlled environment, but folding any laundry anywhere.
> It is not true that whatever a human can do, a machine can be taught to do. The human must have insight into HOW they do it in order to teach it, or otherwise come up with some new algorithm.
Why does this mean a machine cannot be taught to do it?
> There are a large class of things humans do that they don't understand the mechanics behind,
It is not true that whatever a human can do, is the correct way to look at it. The fine motor skills and insight needed for many tasks will always be beyond that of a machine. A machine also has a great deal of trouble with adapting to things popping up that it has to deal with.
> Not true, we just have different priors and are fooled in different ways. See stage magic, optical illusions, etc...
But we can often recognize them as such, which is important. Actually, on that note, is there work on making machines being able to recognize magic tricks?
> But we can often recognize them as such, which is importan
Can we? The occult, new age, and religious sections of bookstores suggests otherwise, as do paid horoscope readings, homeopathic medicines, un/lucky numbers (housing and lottery), and shell-game scams.
All of those (along with stage magic and illusions) have plenty of material (in bookstores and elsewhere) describing either the mechanics, the long odds, or debunking them largely as pseudoscience and/or scams. So it's clear that some people at least can recognize them.
I think it's also possible for humans to be 100% aware of the "adversarial attack", and still use these types of mediums for light entertainment. This seems to describe many people who occasionally buy lottery tickets for entertainment, and would probably apply to many who attend and produce modern stage magic shows.
(In fact, I notice that some of the top modern crusaders against con artists who do use illusions and paranormal / occult claims for adversarial reasons are stage magicians themselves -- James Randi, Penn and Teller, Derren Brown.)
“Some” is not ”often”. If there are any illusions that affect all humans, then by definition we cannot give them as examples because nobody realises they exist.
Yes, it is possible for people to know lottery odds and still play for the excitement. This does not invalidate the claim that many play the lottery with the expectation of winning, nor that people choose numbers superstitiously.
</rant>