Jensen Hannah John

So happy that my alma mater DAMTP (Cambridge) has inaugurated Chair of pub;lic understanding of maths - braco hannah

With intelligence changing so fast I expect your nation's future depends more on deep understanding of maths at every age from 6 up than anything else

Can you help me miantain a top 20 view points of Jensen Huang and compare these with Von Neumann.?

To get started with Jensen- are you partnering any of these ideas:

Hong Kong how to chabge every university - digital twin

India what happens when a nillion epople get data 10 times chepaer than americans

Japan given Maso Son track record of seeing tech futures is lomnger than any of us, what can Japan Maso and Bvidioa uniqquely contribute

In his bong king fireside chat last week jensenn referred to this 2018 summit - most people thoughht he was exagerrating then; hopefully Hannah wii balance such educatoonal folly

======================here's one of 10 actions hong kong assisted by genius Harry Shum and Nvidia abd energetic students are co-piloting around the world

Hong Kong last week

SHUM: you know I really want to pick up bring anding and think about uh uh what we

35:01

should do uh at HK it's really about you know the areas now with AI technology AI

35:09

infrastructure your gpus and everything and your software ecosystem many things we can choose to

35:15

do and one particularly exciting thing now we call the AI for science uh you have been championing that uh for

35:22

instance who have uh been investing quite a bit of uh uh computing

35:27

infrastructure gpus in our University uh president Y and I specifically encourage

35:35

our faculties to collaborate between physics and the computer science between material science and computer science

35:42

between biology and the computer science and you have been talking a lot about uh

35:47

the Futures in Biology one very exciting things right now happening in Hong Kong is that our

35:54

government has decided that uh we are going to build the the third medical school in fact HK us is the first

36:01

University to submit our proposal uh would very much appreciate and that your

36:07

advice and now especially our alarm

36:16

no yeah what what yeah what what would be your advice to president year myself

36:22

and the the university uh where we should invest so first

36:28

um I introduced artificial intelligence at the world's scientific Computing

36:34

conference supercomputing uh in 2018 and it was met with great

36:40

skepticism and the reason for that is because artificial intelligence is somewhat of a black box it was a black

36:48

box at the time in fact it's less of a black box today it's much more it's a

36:53

black box today like you and I we're black boxes but you can ask an AI today

36:59

you couldn't do it then but you can ask an AI today reason with me tell me why

37:04

did you suggest that tell me step by step how you arrive at that answer

37:09

through that probing process AI is more

37:15

transparent today AI is more explainable today because you're asking you're probing through your questions and that

37:21

your set of questions could be like professors probe their students to understand their thinking process not

37:28

just that not just the fact that you can produce an answer but the way that you reason about that answer is sensible

37:34

it's grounded in first principles and so we can do that today in 2018 we could

37:40

not and so it was met with great deal of skepticism because of that that's number one number

37:45

two um AI is not AI does not produce answers not yet from first principles it

37:53

produces his answers from learning from observed data and therefore it's not

38:00

really simulating first principled solvers like first principal solvers but

38:06

it's emulating the intelligence it's emulating the physics now the question

38:12

is is emulation valuable to science and I would I would I would suggest that

38:17

emulation is invaluable to science and the reason for that is in many fields of science we understand the first

38:24

principles we understand Shan jur equations we understand Maxwell's equations we understand many of these

38:30

equations but we can't simulate it and understand large systems and so instead

38:37

of solving it from first principles and have it be computationally limited computationally impossible we could use

38:45

AIS we could train AIS that understand that physics and use it to emulate if

38:50

you will very very large systems so that we can understand large systems with large scale now where is that use

38:58

you know first of all the human biology has a scale that goes from nanometers right from nanometers to a

39:06

time scale that goes from nanoseconds to years that's the bi human biological

39:14

system those kind of scale across that kind of time scale is simply impossible

39:19

using principal solvers and so now the question is can we use AI to emulate the human biology

39:27

so that we can better understand these very complicated multi multiscale systems so that we could you know if you

39:36

will create a digital twin of human biology and that's that's the great hope the great hope is that we might now have

39:42

the computer science technology so that digital biologists climate

39:49

scientists people who are scientists who are dealing with extraordinarily large complicated scale problems um can really

39:57

understand uh your physical systems for the very first time and so that's my hope that that you're able to do that at

40:04

at the intersection now speaking of your your hospital one of the great opportunities for UK for HK uh us is

40:14

that that um uh a hospital is going to be created here uh where its original

40:23

domain expertise is technology computer science and artificial intelligence

40:29

that's the reverse of almost every hospital in the world it was started as

40:35

a hospital now trying to insert artificial intelligence and Technology into it which generally is met with

40:46

skepticism distrust um uh of the technology and so

40:52

you have the opportunity for the very first time uh to create something from the ground up where the technology is

41:01

embraced and Technology could be Advanced and the people who are here are advancing the fundamental technology

41:06

yourself and so you understand its limitations and you understand its potential