all right thanks everyone for for being here uh my name is Ami badani I'm the chief marketing officer at arm for folks that don't know arm we're a uh Central processor for most everything you interact with every connected device um from your smartphones to your Theros smart thermostat in your home to your I'm sure you if you have a smart doorbell to Data Center servers power everything from um AI to uh traditional compute so uh that's a little bit about arm uh I have some esteemed panelists here that I wanted to introduce and you know we wanted to really talk about the intersection of universities and private sector because that's what we think we we really need to accelerate the future of AI um we recently put out or are putting out a uh AI barometer survey where we surveyed approximately um thousands of people global leaders across across the world and about 80% of them have started to apply AI in some form or fashion in their companies and so it's quite remarkable just uh over the last several years post chat gbt how advanced uh AI has become in terms of applicability so we'll talk a lot about uh some of the research that uh the panelists are doing and applicability of AI but let me let me introduce the panelists so we have uh faram who's the president of caran melon University uh I was actually um impressed with the amount of women you have in your engineering program three times the national average is is um what I learned and then we have um Co from KO University and uh your University is one of the Pioneers in Quantum Computing so we'll talk a little bit about that and then we have Eric from I'm I'm going to get it right Muhammad bin zad University of AI which for folks that don't know was a university created in 2019 the first AI University in the world uh so excited to have all three of you guys would love for you guys to do your your own introductions so faram why don't we start with you sure happy to be first of all thank you to arm and thanks to you for uh hosting this panel it's good to be with you and it's good to be with my uh two good friends and and collaborators to engage in this in this panel yeah great well thank you and I'm very uh happy to be here um I'm a quantum computer scientist but at the same time in order to advance Quantum we have to really keep up with what's going on with AI world so I'm happy to be talking to my friends here yeah I want to also want to thank arm you know for having me in this distinguished panel and I'm very humbled to be sitting alongside with my good friend F he has been a mentor you know of me for many years I had had a good fortune to learn a lot from my experience from carneg Milo in creating this new University great yeah as I learned the three of you guys are good friends so it's it's a small world that's why it's a small world as I've learned uh okay so let's start I mean last April there was an announcement with carne melon and KO University uh where you guys Advanced uh a partnersh or you guys announced a partnership to Advanced AI research with folks like ourselves SoftBank uh micro oft and I think it was in the order of $110 billion if I'm not mistaken million sorry $10 million billion is nice no $110 million sorry uh tell us a little bit more about the partnership well first of all Let's uh appreciate the fact that AI is probably one of the most fundamental intellectual developments of our time and the impact of that on every sector of economy is undeniable when it comes to healthcare when it comes to finance manufacturing and the list goes on and certainly it has an impact on education I know we'll talk about that so AI isn't just a tool in our technological Arsenal but a catalyst for potential transformative change but there's also realization that this change and transformation can be Amplified through collaboration of public sector private sector of course academic institutions with the support of a Japanese Japanese government government and the US government uh we came together there are four institutions that are involved uh uh Q University and Carnegie melon are Partners In This plus support uh from a number of private sector companies including arm and including uh Microsoft and and a number of others and SoftBank as you mentioned and the idea really behind it is to bring researchers and practitioners together including our faculty and students to work on complex AI problems and try to advance it and it's just based on observation that by bringing teams with diverse backgrounds together we can push the envelope so that's really the Genesis of it and we got started last year and of course it's incentivized by support from the private sector which will amplify transfer of knowledge to practice but equally important the symbiotic relationship that exists between the public and the private sector can really be catalyzed through this collaboration great how how do you decide what universities what private sectors to partner with when when you so we were asked collaborations to because um our prime minister kishida Was preparing for his State visit to the United States of America we were asked to part possibly think about partnership in AI with one of the universities in in in the US and natural uh Choice was a Carne Millan because our previous Chancellor was teaching at car Milan uh two of the Emeritus professors Emeritus at Ko were also teaching at Carney melan right now we have 17 computer scientists not a big school but still 17 computer scientists and five of them had an experience of Performing research at as posts or you know some sort of a uh research performing research at CMU so you know we have so many uh hisory so much history of working together so so when we I was asked uh I said carne M it was a natural choice and luckily caring millon felt the same way absolutely and of course the support of the private sector is is crucial in in this collaboration without any doubt uh but Ito is absolutely right there was already existing collaboration and a lot of Partnerships the the challenge here was to really bring it together bring some Focus together and focus in certain number of areas such as multimodal and multilingual learning embodied AI autonomous AI symbiosis with humans life sciences and AI it's an area that I know Eric has tremendous amount of expertise in and also the role of AI for scientific discovery so you see these are potentially transformative areas of research that we're just embarking on in collaboration with Q University great great how how do you guys balance um you know universities are sort of at the Forefront of AI research and and Eric would love to hear from you how do you balance long-term research with you know short-term initiatives or short-term promises curious to hear your thoughts yeah thanks for the question it's a very important one always in my mind and in fact the creation of a university of AI in Abu Dhabi is a experiment in the country's Citizen and Leadership to answer that question because UAE is a very young country uh it is heavily relying on you know fossil fuel economy and so forth but uh there is a sense of urgency that this train is leaving you know AI is the new engine of a future economy and technology and uh we don't want to be left behind we want to be in the game not just as a consumer but also as maybe a contributor and the innovator maybe even as a leader so how to make that happen it's not just building you know a R&D arm you in a company you know or importing a lab you know from uh uh you know big corporations we need to have our own you know uh Think Tank and also educational facility to you know start from the ground up starting the culture of St starting the culture of technology and Innovation um in terms of balancing the longterm uh and the the the the the shortterm productivity uh the universe has been thinking very hard we benefit from uh not you know inheriting any Legacy you know from the old practice we are trying to recreate a university program that does both fundamental research but also using that in a fast-tracked fashion to generate productivities for example uh MB become the maybe the few if not the only University in the world that is able to train a large language model on our own facility with our own people right so these are examples about the balancing and the trading of you know uh the risk and also the productivity both from both end yeah great add something oh yeah so in in our case quantum Computing is our long-term research project and uh Quant the status of the quantum Computing right now is like the 60s of the semiconductor chips right so so we are now seeing new mors law kind of a growth of the quantum computers and we're trying to develop algorithm software that will go along with it so that at some point some of the problems that W that will not be able to solve by extension today's computer can be solved by quantum computers and one of the one of the you know good um possibility possibilities is AI in the quantum world so molecules uh when when you actually try to calculate chemical reaction you always have to transfer the molecule exists in the quantum world and we have to transfer such data into today's computer digital world uh and this data transfer takes so much time in fact it grows um exponentially with the data size so this going back and forth between classical and Quantum world is very challenging this is why it's very difficult to calculate uh chemical reaction in using today's computer digital computer but if creating molecules itself within the quantum world is easy because we can just create them but then we can actually if we can perform uh deep learning kind of you know machine learning within the quantum world then we will be in business because you know we can actually identify which molecule is is is is is you know um effective on certain uh far pharmaceutical applications so on so we want to create um you know AI within this Quantum world and this is our long-term research but in order to perform such artificial intelligence research in a Quantum world we need to really understand the state of Art in Ai and this is why we were Clos it together with today's AI people head Cutting Edge AI people and you know try to you know put them together if I can add to that but to answer your question just in the context of emerging Technologies I think what we're seeing is the pace to scope and the scale of these advances is unprecedented and and for that that reason alone it's really transforming the way we approach research uh to the point that Eric made it's actually not really that useful to make a distinction between say basic research and applied research as Eric was alluding to research is a Continuum so you really not only have to focus on fundamental research which may take decades to come to fruition or never but you also have to look at translational research and be able to essentially take some of the foundational work that's being done to practice for that reason alone I think the nature of research is changing in several fundamental ways one is research as you mentioned has become much much more interdisciplinary so to solve some of the most foundational challenges that we face in society whether it's energy whether it's climate whether it's Healthcare uh all of these require essentially interdisciplinary teams come together and AI is an accelerant AI will definitely accelerate research Discovery and this is something the realization that I think many researchers uh have have accepted the second I think uh important factor is beyond interdisciplinary research is we have to recognize that partnership with the private sector and ability essentially to translate research is going to be extremely important because the pace the scope and scale is is so rapid that we need to recognize that in fact this uh the inherent relation ship that exists between the technologies that may be essentially applicable in the private sector today versus the research has to be really explored for that reason I think that changes the final thing I want to say is that there's also a rapid convergence with other uh emerging Technologies for example between Ai and biology between Ai and materials for that reason what we're seeing is in fact again these teams that are coming together are not just bringing a narrow perspective but bringing a much perspective to addressing these problems I I love what you said about changing research has to change with AI accelerating at the pace it it has in order to sort of continue to accelerate you really need to change how research is done and I know at arm we have a a university program and we collaborate with many different universities car G melon included and my former employer actually at Nvidia we did the same thing and and we realized how important it was to collaborate with universities because that's where a lot of The Cutting Edge advancements come from is a lot of universities so that collaboration between universities and P private sector is really critically important for us in terms of how we advance uh the future I think that's spot on and I think U looking down in the next few years I think we're going actually see even uh more changes in the way universities or there should say the private sector interacts and works with the with universities and that's all for good that's actually positive including a number of policy issues that have to be tackled with respect to Ai and emerging technology so that really does require the private sector and the universities and the governments to come together work on some of these issues yeah great so back on on Quantum Computing so obviously Ko's been at the the tip of the spear in terms of uh all the advancements with Quantum Computing at uh CES a couple weeks ago Nvidia recently discussed how we're at least a decade away from some of these practical applications of quantum Computing do you are you seeing it the same way are there things we can do to advance that so so um we started um IBM Quantum Hub in 2018 so we became the first IBM Network Hub on Quantum uh already seven years ago and the first IBM quantum computer chip that we tested could not do anything even a one gate operation wow 70 years ago but that was a newborn chip and after two months two months later the next chip were able to perform only one gate with one cubit but that's you know that was a significant investment so so we've been in a way parenting all the devices chips that are that have been being you know grown by IBM so this interaction of parenting and also chip growing at IBM has been um you know very rewarding experience and you know we've been you know U and and in this regard uh we actually from the very first day we invited partner companies to join us so we started off from two uh Banks and two chemical engineering companies and of course when they first saw the performance of our of their first chip they were devastated you know because it you know chip can do anything but then after that you know we've been witnessing the growth of the Chip And then with this uh people start to feel confident that there is a future but you know companies leave when there's no future so unlike the government project so we've been fortunate to uh have very nice collaboration because they send us they they send their best employees or scientists full-time to K University so we all work together and even number of banks work together and published papers together so that is very new and then um now we have nine companies joining us and no zero company left us so so in the sense that um qu computer scientifically it is very difficult to rigorously uh predict when quantum computer will exceed uh today's computer uh you know uh in in which year and what which problem and so on but at least those companies who are working with us feel confident or or comfortable that there is a future so that they stick to uh they stay with us yeah yeah that's great Eric back to you so you you've obviously done a lot of research in the world of AI models being applied to things like biology uh and you've been a b big advocate of Open Source Frameworks uh I'm curious there's you know a lot of debate about open source versus closed Source models what's your view do you believe that we need to have open Frameworks to advance the state of AI look at Linux right I think uh you know uh I would say 90% if not more including Apple's operating system is built on the inspiration and the structure and design from uh Linux uh open source you know uh is uh not just uh you know a strategy you know it is a philosophy it is how scientists carry out their work now extending to you know the Computing IND and also the research area so I think it is essential to recognize the importance of uh not just open source but really open knowledge open your process yeah you know so that uh we form Community to jointly look at interesting challenging and even sometimes risky problems together because that's the best way to prevent risk and to also mitigate the risk I also want to add that uh people thought right now ai is taking all the word and the where are there already to consume the AI results I I want to add that in fact it's just beginning in of a whole Revolution there are lot of lot of areas that is not yet touched upon by AI or maybe not touched upon by AI or Computing deeply taking biology example you know you may heard that the drug design is everybody's talking about but drug design only occupies maybe less than five or 10% of the whole worklow and what's the remaining is about uh uh trial and the experiments and validations and even efficiencies in and that's a problem because uh if you look at for example in chip design you know how much time is spent on actually the trial and how much effort is spent on the design and the manufacturing there is a different equation there some Fields like biology actually still live in the good age they still use the same method from Antiquity try and the error try your luck and uh uh stir up a few cells and so forth there's a huge opportunity for AI compute to disrupt the status quo you can imagine you know just like what you just said the law of quantum you know in AI Computing AI itself can be the new math you know for many scientific research to allow all possibilities to be computed and simulate in a skill that has never happened before right so I think U as a university we have the freedom and also the responsibility to also do something different from uh industry and uh and Enterprises because uh the role of the university is to learn and also to do knowledge creation so I think uh again that goes back to your uh comment on the open source want to create you know a new culture and a new um you know maybe mechanism that is compatible you know with the amount of Revolution and the speed of Revolution that we are facing on this AI yeah so you know I was going to build on something that Eric said about scientific discovery and and the potential for AI to accelerate scientific discovery at Carnegie melon we're experimenting with something called automated science just imagine a laboratory that exists that it's not sitting next to a faculty member or graduate students where a lot of the wet lab experiments can actually happen remotely and in a automated fashion so like a virtual Factory of exactly in the same way that we can think about software moving in the cloud think of the laboratory the the wet Labs that you do experiments in biology and chemistry and Material Science C with someplace such that it has hundreds of instruments and it's accessible to a broad group of scientists not only potentially at Carnegie melon but at other institutions and now here's the deal when you make that available and you can automate that you can gain significant amount of efficiency in the scientific discovery process now if you bring AI into the picture and you bring machine learning into the equation what really does is the potential is there and we've shown this in a number of cases the potential is there to actually accelerate that by pruning the search space of essentially different experimentation after interesting this could truly turn the scientific discovery process on its head and accelerate in levels that we've never seen before and of course the direct impact of that is going to be in a number of areas including life sciences and Health Sciences that Eric alluded to so there's been a lot of talk I know at we and other other um forums about 2025 being the year of practicality of AI we're going to actually see practical applications in use cases of AI what's your view are you guys seeing what are some practical applications that you guys are seeing of AI in either in your current everyday or things that you've been sparked your interest you guys want to start but go ahead so many different applications happening right now I mean you know you talk about um applications to bioscience uh we uh our medical school has strong uh research uh program on microbiome and you know uh each one of us carry one kilogram of microbiome by average it's it's external entity and you know we have you know tens of trillions of those is you know a microbiome and you know people think that that believe that this is you know understanding of microbiome within our body is one of the keys for for the longetivity so so so but it's a big data and bio bioscientists are just living with huge data they collected without knowing what how they should really tackle um but you know now they're working with our AI scientists and you know they're suddenly discovering all these things you know new new new new Sciences new sciences and of course you know then SCI AI scientists look at their process of collecting data and just like farum said they could just immediately propose that hey you can automate this right in this way so that you know you can really speed up you know collecting data and so on so this kind of a you know um collaboration or discussion is just you know is it's transformative yeah I mean consider the fact that we can now today create text audio video images code uh that we couldn't have done just 5 years ago we can do instant language translation we can Essen uh recognize face and objects we can navigate roads and traffic these are all the cool areas that we talk about all the time but I think the applications that are actually having a huge impact today is how AI is being integrated into mundane routine applications like CRM like into workflow of insurance companies yeah that those are going to be the low hanging fruits and those are going to be the areas that we're going to see significant impact on the on the business business sector various sectors of the economy while we're still exploring accelerating scientific discovery while we're trying to trying to solve some of the most fundamental challenges that Humanity faces but I think the application of AI the new generation of gener generative AI especially uh to routine workflow and business applications is a lwh hanging fruit and I think that's going to come to fruition much faster than we think yeah I I was uh astounded by the survey that I that I mentioned earlier this AI barometer survey over 90% of global leaders have put AI in practice in some way shape or form in their applications which is you know a huge number I I assumed it would be somewhere in the 50% but I was surprised to have it be 90% Eric what about yourself yeah I'm not sure you know uh there is a magical year that application will happen you know or uh get apportation but I do believe that uh the speed of adoption you know of Ai and also the uh the general public reception of AI being part of their life and workflow as F mentioned will be shortened if you go back to history you know how many years uh it takes for the world to adopt electricity 30 years probably and how many years it take the world to adopt automobile I just heard a story that you know uh uh you know UK actually British know started this whole Innovation but why Hy for took it to the US it's because in UK they were basically this allow car to drive over the speed of five miles hour yeah so there are interesting lessons that I hope our policy makers and the general public can draw experience from that uh you know facing new technology and the disruption what is the best way the graceful but also productive way of uh integrating them and welcoming them but also regulating them in a balanced fashion so I think that would be probably the most interesting in topic among the general public and also the scientists yeah you know uh you know in the next few years and University actually plays a very unique role you know in this because we are both the inventor of Technology but also the educator of uh this technology right we interact with not the science the natural world but also interact with people so we have the unique opportunity to be the advocate and of this new technology to really set a course right and also uh get the m correct right now lot of BU there a lot of hype lot of misunderstanding about AI that actually bothered me the most because I getting frustrated that people were talking AI in a way that I see very differently so that could be a a interesting you know uh new opportunity for the Educators La last question for you guys uh so according to we there's going to be 170 million new jobs created by 2030 I think we can probably all agree a lot of those will be in some form or fashion whether you apply it to those jobs or whether they will be integral in terms of you know the new job creation what are you guys doing to prepare your students for sort of this world going forward quite a bit as it turns out so a couple of thoughts on this we are actually reimagining curriculum uh not only the issue of what do we teach our students when it comes to foundational AI but also looking at other disciplines and looking out a curriculum and see how AI is going to impact that case in point our College of Engineering for example at carigi melon recently launched a series of Masters in Ai and Engineering application of AI to civil engineering mechanical engineering Robotics and so on and so on so that's that's a new way of thinking about curriculum and how we approach it the second U issue the second point that I want to make is that I think we need to recognize that everything has become interdisciplinary right so as I mentioned and a number of others alluded to this is that to solve these massive problems we have in society we need to approach it with a interdisciplinary lens for that reason not only we need to teach students who have disciplinary expertise but equally important they can connect to other disciplines that has become more important than ever before the final thing I want to say is the role of Technology itself in transforming education and often people talk about you know if you're my age every 10 years somebody comes in with a new technology and says it revolutionizes education so 30 years ago maybe it was laptop and then if you give every kid a laptop education will be transformed then it was the internet if you give everybody connectivity well you know it has had an effect and in fact 10 years ago it was muks people thought muks will trans essentially transform education the truth is that these Technologies help but fundamentally the question is how how can technology improve the learning outcomes that's really at the heart of it yeah and I think today what we're looking at is the role of AI and role of data and computation in improving learning outcome for our students the the way to really think about it is that there's tons of data that shows that if a student has a tutor if you give each student a tutor it improves learning outcomes yeah so the question is can we use technology can we use AI to make education to be much more personalized much more adaptive and geared toward the progress that the students makes so it's not only bringing technology to improving learning outcomes but you actually have to change the curriculum such that as a students studying it you know the underlying system is monitoring it constantly evaluating what the student is learning and using that to in fact change the curriculum yeah what the student is learning as it makes progress so ultimately what we're going to give students whether it's K through 12 or higher ed their own tutor when it comes to learning I think that is sort of the dream about education that's what we aspire to personalized education absolutely and much more adaptive yeah so just in general students are much better than professors in adapting new technologies especially you know for AI tools um professors are too busy or and and and what I found out in 2019 so that was already six years ago I start we started so-called AI Consortium in which we have open call for teachers among uh on AI and those teachers are going to be basically our students so we ask our we we we collect good students in AI to teach other students right that's great and uh that's been going on for past six years and now with the emergence of J they're now teaching faculty members wow and you know faculty members you know put out all these homeworks and sometimes they point out that you know your homework is somewhat useless because our new tools can actually do this in five minutes or whatever right so so so it's this kind of a new interaction between students and faculty members we found very useful of course you know some sometimes it's intimidating to some of the professors because they really have to change the way they teach right and they don't want to hear that their way of teaching is becoming oldfashioned but to tell I mean just to tell the them the truth I mean this is what's happening so uh our AI Consortium where good students teach other students and good students each other are faculty members I think it's beginning to get some meaning and actually is getting recognized as a good platform I love that I like the I I love the flywheel Eric anything any last minute thoughts from you I cannot agree with what has been said you're brilliant you know Educators uh on the on the panel so for us as a Young University I to say that we were blessed with all this huge B of knowledge that we can learn from and uh and also the lessons learned in the past but also as a Young University it's also a place to experiment and to try some new ideas because we don't have the baggage you don't uh need to pay the cost of disrupting the Legacy and so forth one thing that come to my mind is uh in addition to uh focusing on the effect of learning how to make learning better and personalized but also to rethink about what to teach if you look at education you know hundreds of years ago you know a calligraphy is something to teach you very important for people to advance their career and then in my age you know we were told taught how to break the square root of two and something like that and now probably very few kids need to learn that right so the body of knowledge are becoming more and more accessible to Common citizen do they need to spend the amount of time they did before to learn that or learn something new in terms of how to sit on those knowledge that is at your fingertips and become more productive this is something I've been thinking with also my colleagues and also uh also the the the the consumers who are the consumers or the people who hire students you know uh into their workflow they are the consumer of the university and uh I often find that the students from the University still need to take one year or two to kind of become fit you know for a Enterprise environment or in a government environment so I think there's opportunity also that AI in the context and the new needs and demand also being becoming obvious and how to as F put it rethink and reinvent curriculum and also the way to teach that great I know we're out of time but I want to build something that Eric said I think it's really really important you know if you think about a kid who's in elementary school today very likely that young boy or young girl is going to end up having jobs that have not even been invented they don't even exist today yeah that's very likely scenario just think about 20 years ago versus today in fact Harvard Business Review had did a study that said halflife of skill is now less than 5 years in for University president is the Same by the way I was going to say this I heard about this but but in the tech Fields actually it's dip to as low as 2.5 years so the point that Eric was making is so important is that it's not about just teaching students a particular skills we should teach them some skills of course but really it's much more foundational we need to teach them how to learn we need to teach him foundational skills we need to teach him also not just Tech technical skills we need to teach them collaboration we need to teach them communication skills a lot of the skills that Humanity almost takes it for granted it is going to be even more important as this rapid pace of Technology transforms every sector of our economy so almost like a reinvention of how we think about work today um I that's how I how I hear you guys talking about it which will be very interesting to see over the the next decade or so uh thank you guys I know you know we all kind of share the same interest of democratizing AI and making it accessible for everyone whether it's at the private sector at a company like arm or across uh the university sector so thank you so much would' love to come back next year together and talk about how we're advancing progress thank you for having us thank you thank you thanks very much you thank you so much thank you so much
2025-01-24 01:49