The Law of Large Numbers and Fat Tailed Distributions 

My inspiration! I teach undergrad Math Stats recitation at NYU Courant, and I always make sure to talk about power laws and show them a simulation like what you showed at 13:36 ... it blows my mind that nobody explains this stuff when they teach basic statistics!!

It's quite interesting to realize these ideas aren't taking into account human behavior. They show that we would need to approach risk differently even if everyone were ethical and transparent. Thanks for your work Mr. Taleb!

After reading your books that includes maybe 8 times and still not fully understanding everything I find this video extremely helpful and supplemental. You should make more supplemental videos

Unbelievable! Great honor to know you Nassim, you lighted my darkness of randomness! Best wishes from Iran...👌🙏

It took me 2 years to understand only 25% of the content described in this video and I am watching it again to understand more. And my education is postdoctoral status in STEM. Truly ahead of time content.

I think audio quality would have a concave curve. Quality is maximum when the needle is at reference deflection of the VU meter, in other words 0 VU. The left side of the curve suffers from poor signal to noise ratio and the right side suffers from distortion.

This guy is So fucking intelligent.... Respect from me

Watching this for 10th time in a row, will probably need 10 more to completely understand this small facet of anti fragility!

Thank you. Very informative. If you could turn up the volume of your micro, that would be wonderful. 😊

Hi, I find the audio for this video segment to be barely intelligible and I think I know what the problem is. I'd guess the audio was recorded using a webcam. Now some webcam software does a form of processing on the audio in order to try to eliminate background noise. Unfortunately if this processing is done too aggressively or if the sounds being recorded are too faint, what will happen is that quieter sounds are either made even more quiet or cut off entirely. Either way the end result is recorded speech that cannot be easily understood. The solution is to disable this processing. If you are using a logitech webcam I'd suggest disabling the RightSound option in the logitech webcam controller. When you do that you will notice that audio recorded will have a background hiss - that is o.k. because it means low level signals will not be cut off and as a result soft speech will actually become more intelligible again. For really good audio a better bet is to buy a USB based microphone of decent quality and use that instead. You may also want to do some dynamic range compression on the recording afterwards as well as some slight equalization boosts if necessary. But that will require more complicated processing software.

Thanks sir , that was very meaningful to my and my current work

barely audible but still thousands of times worth watch

If I get the essence, can we trust or to which extent we can trust a random sample of a fat tail distribution to make sensible inferences?

I just came across this. Not a complete idiot but I'm trying to work my way up to a gut feel, you look at it an know. Thx for this!

Why is there a soda in the middle of probability distribution slides.

Thanks.

Excellent. I was wondering what are the implications of this in econometric modeling (especially macro-econometric models).

I love how he said Physicist care about Power laws they prefer to be clean and do not care about stats consequences of the fat tails. I'm reading his current statistical consequences in fat tails now.

Yeah. Needs to be louder.

لماذا لا تحاضر باللغة العربية، دكتور؟

It would be nice to have the slides Ill screenshot them for now I guess :D

Thank you for sharing.

The video is great, but so quite...

Modern way of giving a seminar

Don t be afraid to take a sip of water from time to time