A conversation between Nassim Nicholas Taleb and Stephen Wolfram at the Wolfram Summer School 2021
Vložit
- čas přidán 3. 05. 2024
- Stephen Wolfram plays the role of Salonnière in this new, on-going series of intellectual explorations with special guests. Watch all of the conversations here: wolfr.am/youtube-sw-conversat...
Follow us on our official social media channels.
Twitter: / wolframresearch
Facebook: / wolframresearch
Instagram: / wolframresearch
LinkedIn: / wolfram-research
Contribute to the official Wolfram Community: community.wolfram.com/
Stay up-to-date on the latest interest at Wolfram Research through our blog: blog.wolfram.com/
Follow Stephen Wolfram's life, interests, and what makes him tick on his blog: writings.stephenwolfram.com/ - Zábava
the stuff I wake up to after falling asleep on youtube is funny as hell
🤣🤣🤣
That’s what just happened to me 😂
Me too!
Same 😅
Dude i'm on acid and i didn't realize the music stopped
Now we're talking about what i can only assume is some sort of time travel algorithm or something?
This made me realize just how good Stephen Wolfram is at explaining things
He's up there with the greats.
Cc ccccccj
Summary:
Stephen Wolfram: "Let's make a physical theory of economics!"
Nassim Taleb: "That's precisely the mistake people make"
For what it's worth, I agree with both: Yes, let's make one; and no, don't ever bet all your savings on any theory.
What a wonderful experience to see two of my heroes together! Please, guys, do more conversations!
41:44 - 43:58 - I think this sums up Nassim Taleb's point on the first hour or so of this video.
Two reasons why above can't happen:
1) 44:11-44:52
2)44:53-46:42
you need more data points in order to get something deterministic, but obtaining more data points causes spuriousness more quickly than the law of averages to take effect.
It's good that someone is attempting to put an equation to economics...but I don't think it's the first time that's happened (I'm not a historian, and have barely a layman's understanding of statistics). I also think Nassim Taleb's arguments are more valid, than the questions.
I think a decent common ground would be, "let's develop a New Kind of Physics to deal with economics." I feel like, "anti-fragility" is in many ways an expression of, "A New Kind Of Science" but in, "practical" terms. Cheers.
In regard to economic numerators and home currencies-- it's kind of like the Speed of Light in General Relatively. Just a thought.
To me, this is a false dichotomy. Taleb AFAIK is pointing out where the standard models of finance fail. Wolfram AFAIK has come up with a computational theory that explains the how/why of the fundamentals of physics. The shallow conclusion is “Wolfram thinks he can solve finance with a predictable model to make traders millions and Taleb says that can never happen”. The more nuanced conclusion is that Wolfram IMHO may be able to explain why the standard models of finance are failing and possibly why the old “high-level rules” are bad approximations of more fundamental quanta that have some transiently stable states or dynamic local minima that they stand models predict. Or something. YMMV 😂 Certainly Wolfram isn't trying to disprove Taleb he is more likely IMHO to vindicate Talab by showing why outliers and black swans are more probably than believed and that only with a perfect simulation of everyones behaviours could you predict them (his theory of computational irreducibility).
Holy fucking shit, that was the best demonstration of the central limit theorem I have ever seen
Time stamp?
If you point out a single thing in a 3.5 hour convo then it is kinda useless without a TS.
17:00 I think
@@laggan 15:25 "Let me explain the central limit using Mathematica..." For future reference, CZcams added the ability to control+F the subtitles on some videos now--super cool feature. It's in the hamburger menu next to the save button, click "Open Transcript", and cntrl+F central limit and you'll find he also mentions it later in the video at 26:20, 43:00, 1:44:36
@@ehhhhhhhhhh you son of a bitch...
Thank you holy whoa this is bigs
@@HopDubstep When you add two uncorrelated random variables, the distribution of the result is the *convolution* of the two variables' distributions. You might also say their characteristic functions are multiplied, and that multiplication in the frequency domain corresponds to convolution in the spatial domain.
Taleb and Wolfram flexing their libraries on each other
That’s a very small sample of their libraries
What a waste of time!! The guy maybe knows options trading/pricing but has no clue about physics, economics or statistical mechanics and doesn’t answer any of the questions but confusing everything with his simple example
@@scoreprinceton I've watched this video about five times over and I agree with you on the fact that he is poor at teaching. After watching this video many times I see the genius though
@@JeremX17 I stopped watching after the first 45 minutes!! Maybe there is something worthwhile - thanks for speaking out
Hahahaha
Respect Wolfram, despite your great wealth you still patrioticly show British dentistry.
LMAOOOOOO
British humour at its finest. Literally made my day.
Golden comment
2:10:50 I meant psychology not statistics.
I love how it feel Stephen cares for us the audience so we don't loose track and explains it fast enough and simple enough for me to understand and go along.
he is a 180 IQ wunderkind and I appreciate him lowering his IQ to our level so we can try to grasp what it's being discussed
it's really just the opposite, Nassim doesn't give a fuck about making himself understandable, he says things quickly without explaining much and just assumes you should do your homework or read his mind or something.
@@TheXV22 He didn't write that Stephen wasn't clear or articulate enough. And about Nassim, well, even if he's behind him intellectually and as a verbal communicator, he isn't talking about complex/difficult stuff for people knowing/understanding the minimum about the topics of their conversation.
I’ll probably understand 10% of this.
That’s generous for me 🥴🥴
If they understood it, they would not be talking about it. It's not a "science."
Slow down the video, stop at words you do not understand and look them up. That should get you to 80%, and that's good.
Such a gift to be able to listen to these two. Thank you both
This conversation is what I envisioned about the future of the internet. I love seeing experts pairing off with other experts, sharing ideas, and actually working through things. We're able to witness the actual process of discovery and the generation of new ideas.
Absolutely phenomenal conversation! Meeting of Math & Markets!
Thank you for posting this! It was delightful!
1:53 "a parameter is right on average, but not as a variable. a function of average is not average of function"
42:57 "law of large number is aggregation, what is central limit theorem is aggregation. something aggregates very well, you don't need to know the detail very well, because the overall is going to work out very well. you build portfolio that works very well on the Gaussian world. the problem is that we don't live in that world."
54:51 "our explanation of time dilation in modern times is the universe is computational. you can either use your computation to compute what happens next in time or you can use their computation to move in space. if you use your computation to move in space, then you will have less computation to use to evolve in time, so time will effectively run slower for you.
1:17:18 "if a currency is just there for illegal activities, in other words the currency for fraudsters ransomware and pariah, cannot be a currency. it needs to have a bunch of suckers around them to trade their currency for it to have a value."
31:20 ”The problem with economics is that we don’t have anything bounding a number”
Tnx for highlights
intuitively, options pricing is similar to time dilation computations
Excellent. Science is a hypothesis under constant revision. Very good discussion and rigorous questioning.
Fantastic watching Mr Taleb tinkering in Mathematica.... such a great way to learn.
You two are amazing , well received.
@ 2:03:30 "N of 1 is a clinical consideration... N of 100 becomes a statistical problem. N of humanity - a pandemic - is a tail risk problem. And the three are three different disciplines" Pure, classic Taleb.
Two of my favourite thinkers, what a find!
This talk was like attending a course,such was the depth and breadth of topics and concepts.
I find it very reassuring to hear mild-mannered, careful Stephen Wolfram talking about religious scientism in public. Fantastic.
What timestamp do they talk religion?
Please do one of these with Ed Thorpe if possible, I wish there were more videos of his talks. He seems very easy to talk to, his pedigree is second to none, and his ability to articulate thoughts is incredible...
The complete opposite of Mr. Taleb! Though it is fair to say, as someone else pointed out, Mr. Taleb is more comfortably speaking in a more technical format.
When I'm thinking statistically (terms of probabilities) I translate my thoughts in terms puts, calls and the accompanying greeks, so I'm right there will Taleb on his first example. Great stuff.
This is amazing to watch and listen to, in real time. Wow
Wolfram said most of the economy is simply people buying and selling things like lettuce. The complexities of that simple transaction, the risk, the risk mitigation strategies from the farmer to the transportation to the retailer is extraordinarily complex, and you would run into the situation that Taleb describes where there isn't enough data or time to calculate a mean.
Taleb's basic point is that you can't model this stuff, and that is where economists fall short. They can observe, see patterns. But they cannot understand the minute workings to build a model that is predictive.
And that there might be a complex explanation or estimate, for lettuce, but you’re now a lettuce expert, not a market expert, and it doesn’t generalize.
Well said. Many financial economists are just pseudo statisticians who dont understand that the applied prob and stat methods they are trying to use dont apply at all to their questions and problems, especially when it comes to modeling, forecasting, and pricing. They try so hard to justify the use of complicated forecasting and modeling techniques when the real issue is that such a task is in and of itself impossible. I think the Austrian school is closest to reality simply by understanding the inherent limits of the economics field itself.
Also, I find it perplexing why so many people, Including wolfram here himself, seem to inadvertently associate taleb and his ideas as some sort of collection of revolutionary "economic" ideas when in reality, like he has said so many times himself, one of his many claims to fame is exposing the academic economics profession for incorrectly using and applying a set of tools from a completely different domain of expertise(statistics), to problems which in reality, may not even be answerable in any closed, elegant form constituting any sort of rigorous theory (derivative pricing and financial forecasting). He is not an economist so why would he be suited to ask for creating any sort of new economic theory. He is a statistician and practicing trader. He thus would be the first to caution against any sort of grandiose economic theory that steps way out of line with reality
I understand only some parts of talks here by two great thinkers I admire. However, I understand at least the importance of their conversation. Good to see such creative and intelligent argumentations.
Dude, sadly, you have to be bald to understand it in it's entirety. I guess it finally payed off to lose all that hair ... .
@@muzzletov uh what? What they’re referring to isn’t even that complicated. It’s a lot of philosophizing and expression of opinion.
@@MrEo89 uh, what? a lot of philosophizing and expression of opinion? Is that your summary? "They used words to describe intent and meaning". What is the rest about then? If you can't summarize it well, or at all, I suppose, you don't understand it in it's "entirety". Also, why would you not be able to understand it without being bald? If it clearly can be seen as a joke, why do you then get offended?
How do you know it's great if you don't understand most of it?
I can feel the energy from the fusion of these two heads
Merci pour le partage !!!
from great complexity, it becomes simple all of a sudden. Well done.
Awesome 👌🏻 especially the discussion on medicine 👍🏻
Amazing long form Education. Paying it forward.
Professor Taleb's analogy about value discrepancies in pricing for particular items is essentially an analogy about energy. In this analogy, one currency has more 'potential' than the other, and the potential equilibrates gradually which is an entropic process. The issue is: we wonder if we ever reach equilibrium in this process. That is the flaw with modern economics: which assumes economic systems are in equilibrium. This is used to justify the use of a gaussian as a starting point for analysis. However, the fact that power law or fat-tailed distributions are so prevalent are strong evidence that the economy does not operate at equilibrium.
In the economy people's desires/preferences create potential differences and desires can change arbitrarily. Not only that but people are averse to uniformity.
This is pure gold.
I'm here listening to you mr.Nassim because of my love of my life adores your trinkile of thoughts. and know i'm sleepless, thanx
I'm disappointed Taleb didn't greet us with his usual "Hello Friends" ;) (great conversation btw)
This is pure gold!
Digital Gold
Very nice, really interesting! I have two questions maybe some of you can answer them :) :
- So what do you do if you need to use p-values at 2:02:36 they kind of conclude what to do, but it is difficult to understand/hear what they say.. My take is, just like physics (5-sigma) and check assumptions ofc.
- Anyone has the link on what Taleb discusses around 2:30:55 here he says Wolfram, his son and Diego computed a model in an afternoon which is fastly more robut compared to traditional epidemiological models. Is there a link of what he refers to?
Thanks!
Random trip to CZcams and what do I find. Two most interesting people whom I follow and admire. Hopefully their intelligence will rub off on me.
Thank you so much
I cannot believe Wolfram has only 85K subscribers. 😮 and this interview only 152K views.
This is effing epic.
My brain is hurting, but in a good way.
I pretty much didn't understand much but it was really fun to listen to.
great talk!!!
Could listen to these two for a while.
Please, can you guys do this every ones in a while? Such a good chemistry between the two greatest of the greatest. May god keep you both safe and happy
Their chemistry is amazing and surprising! Particularly because a lot of people /"experts" find both of them to be difficult personalities lol
brilliant topic! wonderful, thank you so much, I was wondering why almost nobody in risk management and even in statistics! never talked about these math backgrounds of statistics and the effects on the real complex world!?
We need one more such conversation between these two.. it’s been a while
Such an important conversation...
Take the average option. That is how it's done correctly! Thank you Nassim Taleb!
No, take two options whose average is 0 but with high upside and low downside
Nicholas Taleb was spot-on with expected future value, but it's surprising nobody mentioned labor costs. It's why a hand-crafted item costs 100x more, and defines a price basis for everything traded.
Cost and price aren’t necessarily equal.
@@cyberft but they can't remain unequal for very long. Their difference comes from forecasting error (or hidden costs / prices).
@@frobnitzem forecasting error is one explanation. You also have other effects like seen in luxury goods, where price far exceeds costs, this can persist through time and may actually become more entrenched due to prestige and branding.
@@cyberft that's just the hidden cost of researching item quality. When you get a brand name item you expect it to have a higher standard (whether its true or not...).
Instead of cost, he brought up the concept of “value” as a subjective quantity and “price” as a public quantity, and presumably there could be value upstream of the end-user. To someone that paid the cost of production, maybe these are highly correlated.
This is absolutely wonderful. I would love to see the next episode. One comment around 30 minutes: there is an analogy between atoms and currency in which energy is suggested but Professor Wolfram correlates it with observability. From the economics side, would information make the most sense, since we are talking price?
No
This is a fascinating conversation. I was looking for motivation to learn statistics, and I think I’ve found it. Most of the examples in basic courses are not very interesting, but it’s encouraging to hear the wide range of applications discussed here. Some of the comments on misapplication remind me of Richard Feynman’s speech on Cargo Cult Science, where people use the language of science, but are not actually applying the science. In my previous jobs, I’ve often been called upon to apply statistics incorrectly when I only new enough about it to know the application was incorrect (such as analyzing results on a sample of ten or less units).
I agree, same with me
Very very good discussion! I didnt understand it all but its awesome
Biggest Taleb intellectual contribution beyond his unique expertise in risk, is his Fragility spectrum concept. All things can be put into these categories: Fragile, Antifragile, Robust. Fragile things are usually made by man, like communism. Antifragile by evolution or huge number of co-creators. Example: society. Robust - are eternal, designed by existence itself, like gold, space, time, gravity. I can expand fragility spectrum into theory of everything which is more universal than any other, incl CTMU by Chris Langan.
At around 3.23.20, Taleb: "I have to be in a restaurant at 7pm", lol.
Commenting for the algorithm. Show me more of this in the future.
this played while I slept and made my phone die
Best video on CZcams - been waiting over a decade for something decent to arrive on the internet. Just 2 years late :)
I realized i feel something when i read Wolfram's writings and I feel something else when I read Nassim's. The same is true when I listen to them. Both are stimulating but in a different way, hard to put in words. Wolfram's thought is linear and at first it feels it easy but i soon realize the point he tries to make has depth which i don't fully understand. Nassim switches b/w examples to make a point and when he finally restates the point("you see"...), I kinda find it hard to associate it with the several examples and lost. I think i need to study more math to understand both.
NNT is a practioner . So ,so much examples
I have a similar experience... and while watching this I had a realization that when I read or listen to Nassim, I rarely if ever have an "Aha!" moment. The times there is an "Aha!" - it tends to be insignificant, like an "Aha - I figured out which concept I am already familiar with he is applying here." Its as if the core concept is never elucidated in conceptual terms. Or perhaps, it's as if he lacks awareness of what a naive-to-his-views listener actually requires in order to no longer be naive to his views. Perhaps its a difference of learning and communication styles, but at the same time I know I'm not the only person to find him.. cryptic is perhaps a fitting word? At worst, I could imagine this being obfuscation - but I don't level that claim at him and truly don't think that's what's being done. A final thought - I guess my hangup could also simply be outsized expectations of the profundity of the things he shares, so that I'm always looking for some nugget of insight that is more novel or enlightening than what actually exists to be found. This is not to disparage his work or thoughts. Amongst his books and writings online, I almost invariably enjoy them on net, and certainly also find thought provoking angles. Just nothing ever truly paradigm shifting in terms of how I see things. Idk. I like to think I'm simply lacking the scaffolding to appreciate, and that one day when I've got all the tools, I'll see the powerful insights that his bravado implies.
@@waldiniman the only "aha" moment he provided me with is that 99% of statistical methods are useless or even harmful when heavy tails are involved - which to be fair is a pretty big "aha". I agree with you that he rarely distills something into a principle, but rather waffles on a lot. I'm also trying to glean some nuggets from his numerous examples, but often he also talks about things he has no clue about, so I'm close to "unfollowing"
Sounds like neither are good at explaining things?
@@ReddSpark Yeah they're better writers both.
This is great stuff.
More content like that please 🙏
I feel like after a certain point these guys were having separate conversations with each other.
Wolfram is good at channelling the unfiltered raw Taleb, into a more accessible format.
What that Taleb is saying Mathematically reduces to this:
f(g(X)) != g(f(X))
The order in which you apply functions/operations to X matters.
Exactly, thats the TLDW
"order matters" is actually shorter than that ))
@@Acid31337 And a link to wikipedia is a better indexer than mere platitudes.
en.wikipedia.org/wiki/Order_theory
Composition of functions is usually not commutative. I'm not sure what that relating to volatility was so important though.
Feynman: "There are 10^11 stars in the galaxy. That used to be a huge number. But it’s only a hundred billion. It’s less than the national deficit! We used to call them astronomical numbers. Now we should call them economical numbers." I think Feynman was saying, 'if you think numbers from the science of astronomy are large, try numbers from economics.' Economics is composed of the aggregated actions of IRRATIONAL PEOPLE (with no constraint) while physics is not. There are limits in physics that do not exist in economics. Physics exists in Mediocristan while economics (and most social matters) exists in Extremistan. They are different worlds with different rules.
Great, very, very informative and eye-opening :)
BTW 3:22:00 Can we say that Valuation - Price relation may be similar to Mass - Weight?
This is awesome
Thanks you
Great conversation hitting on several ideas that became relevant to me recently, awesome!
Also 1:00:05 made me lol
That was pretty damn good.
The thing with NFTs is that people don't actually buy them because they're fun. They buy them because they want to make a profit.
"It's gotta be bitcoin all the way down" hahaha. (Lex would approve)
Because people have different values, they can create value for one another through a trade in a market. Markets are where people go to "value-up."
I'm from KY USA & understand him well..
Nice, very nice conversation lots of respect. I used to love mathematica back in school. If you don't want to compromise the statistics but can't afford it anymore try julia.
Amazing seeing these 2 giants together.
Very interesting, was on and off on it in bed. There was a slight interruption in the lingo, when it came to curvature of space-time. It s a "simple" formalism.
There seems to be a way to steal from relativity towards finance. This is a vision tho, digging is necessary.
2 great thinkers . One straight as a straight line . Other connecting dots
Wonderful
Just found out he worked on the movie "Arrival". That's so cool!
Basically, the behavior of unbounded sets is computationally irreducible.
Great discussion. I suspect that Stephen and the team might unpick the rules that govern the economy by interviewing James Harris Simons the founder of Renaissance Technologies. Having all three polymaths in one meeting would be a dream.
Darryl Anka exceeds all three of these gents in spectrum of understanding
One basic problem with this conversación is that Taleb is not an economist and Wolfram wants to talks about the very basic axioms of económics that I presume Taleb does not know (and also does not care)
This is so cool. He’s really good at explaining things.
stephen wolfram is maybe the ideal conversation partner for nassim
Was thinking the exact thing.
Wolfram shows great patience. He listens for 30 minutes as Taleb explains “the mean” and Wolfram is like “ok” every few minutes. Only three more hours of listening to go on this I’m hoping there’s something good. Where’s the beef? I was hoping they would get on to the subjective nature of “implied volatility” I was hoping Wolfram could tear it apart. Anyways unfortunately I can barely understand this stuff I wish someone would summarize it in a few paragraphs.
I fell asleep and woke up to this video
Same here
Bravo!
This was awesome :)
It feels like reading Leibniz to Bernoulli correspondence, but in real time.
On the intrinsic value of Gold, it's industrial utility is the same as silver and as jewelery taste has changed silver use is has increased with new generations, so by that metric gold should be valued around $25 not $2500.
Or silver should be at $2000 ;)
that was one of the quickest shirt changing I have ever seen
Awesome. Simplicity is the survivor.
Holy crap I was 2 hours in and it really starts to cook, and I thought, "aww I hope this isn't going to end soon"... Nope, over an hour to go
At approx 39 mins Stephen Wolfram indicates that there is no axiomatic theory of economics. Unfortunately, this is mistaken. Around the first five chapters of every advanced microeconomics textbook are devoted to proving these axioms. Whether the axiomatic approach is a sound approach to economic behaviour is sound, is another topic. But it exists.
I don’t have economic background but axioms are per definition unproven.
Can't wait for SW digestion of Nassim's scattered ideas
A physicist, virologist and quant go into a bar and each want to get the best cocktail for the best price. The physicist tries every cocktail and says Manhattan is best. The quant looks at all the sales books and goes with the Mojito. The MD, says “You guys are morons. The scientific method is supreme. Hey, bartender, what do you think I should get?”
I don’t get it
It means they all think they are the right and competing over the options when each option has merits
His idea of the scientific method is to ask an authority. @@GodGuy8
Science seems to be a commonly misunderstood concept because people have such a polarized view of it and it should be the most neutral in substance. For the most part, what I get a lot these days is "just cause 'science' says it's true, doesn't mean it is". On the other hand, you get those people who religiously follow "science" and if it goes against what "science" says, it must be false. It is odd that people refer to it as "science says". Science is not a person or group of people. Science is not a place or a thing. Science is merely discovery. The discovery of what is. Learning how to understand our reality. Those answers change all the time but there are also many things that remain constant. Science is not in the theory. It is in the discovery. Discovery can only ever hold the value of true. If "discovery" holds false, it was not a discovery. Science is perceived and theory is conveyed. As we know already, translations are not always correct. So, it is weird to see how many people view it in such a polarizing light of good or bad, right or wrong. It only ever just 'is'. Science is abstract. It cannot 'say' anything. It only 'does'. Scientists can convey their theory based on their discoveries but, theories are ever changing based on new discoveries. No use in putting an umbrella term on "science" like it is some union of like-minded people. It is quite the opposite.
Lol...this was actually very interesting...I opened mathematica and followed through.
I was an atheist until I found about about Stephen Wolfram.
?
This has to be a joke to say he is a god....? 😂
@@simonmassey8850 I mean, I now believe in Goddess Namagiri who supposedly gave Ramanujan the correct mathematical results he claimed came from her. Stephen just made me appreciate this fact. 😉
I just read Wolfram’s biography and I feel like giving up all my studies and living a simple life.
which is his biography?
In chemistry you deal with very large numbers.
Even high school students know 6.02 x 10^23.
There is no equivalently large numbers in economics.