tv Hearing on Chinese Education Emerging Industries CSPAN March 11, 2023 5:02am-6:16am EST
labs because this is an education and training hearin i will look forward to questions for th record on those issues. with that, i think we will take a 30 mute break and be back for the final panel. it is in the schedule, how long we are allowed. thank, it was helpful, informative. i learned a great al. thank you all. >> [indiscerble conversatis]
>> dr. simon wl talk about china's promotion of the semiconductor industries's master witnesses to the remarks to seven minutes to serve time for russians and answers. we will begin with ms. peterson. >> cochairand distinguish mmissioners and staff, thank you for the opportunity to testify on how china's education system is training its artificial intelligence workforce. today, i will assess h china has plans to put talent at its core and shape its ai education system. i will conclude with four recommendations with how the u.s. cannick results on ai talent competitiveness. china's well aware of the importance of ai talent in fueling its ambition and has noted that ihas key talent shortages. in 2020, china's ministry of human resources and social security released a report that
quantified and ai talent gap of more than 5 million workers. only one person for every 10 openings for ai engineers and technicians. that report stated that without progress, thealent gap could exceed tell million by 2025. -- 10 millio by 2025. china has relied on several ai education specific goverent plans which i detail in my written testimony. notably, china has taken a holistic approach to ai education and its national ai strategy defines ai education with the inherently interdisciplinary ai plus x model which integrates ai and other subjects and that includes technical topics such as math, computer science and biology but also explicitly includes sociology and law. as a direct response to these plans, the ministry of education mandated ai in high school
curricula in 2017 and standardized and undergraduate ai major in 2019. this is now offered at over 440 unersities and ichina's mos polar new major. universities have also launched more than 50 ai institutes. this exceeds the plans goals to have at least 100 ai majors and 50 ai institutes by 2020. many of china's most elite universities offer the ai major including all of the ivleague equivalency and all of the seven sons of national defense. 2022 diverges from this elite track. with a few exceptions such as peking university, the 95 institutions that added the ai major in 2022 were mainly lower tier including multiple vocational colleges in rural areas. while this could expand china's ai talent pipeline, it also runs
the risk that china's push could lead to limited instructional resources. this is especially pronounced in under resourced areas which could produce underwhelming results. beyond the undergraduate level, another strategic plan calls for increasing the number of graduate students studying ai especially at the doctoral level. this plan also highly encourages industry partnerships. it asks leading ai companies to train university instructors and the latest cutting edge methods and for company rearch ito ve doue emplment a universities and r&d labs and certification training and host entrepreneurship. to cite one example of industry partnership, one company that's on the u.s. entity list for human rights violations has
partnered with public colleges and universities to launch several ai colleges across china. students are able to work directly with engineers on their voice recognition and translation projects. i will close by examining implications. since the u.s. education system is highly decentralized, it has a much less uniform approach -- approach to ai education than china. the u. is currently prioriting computer science educion mo than . well this an esntial component of training talent to potentially perform ai work, china has shown is a step ahead through its incorporation of cs education into formalized ai study programs. these programs are designed to include some of the latest cutting edge ai research funng and are directly informed by partnerships with leading chinese ai companies. furthermore, resrch has shown
that despite china's two decades of talent recruitment drives, nationals either do not return or do so part-time mostly due to weariness of working in china and potentially confronting workplace politics. a recent scien study found that for those who return to the talent plan, recruits are are of high calendar and show a productivity on due to larger access to funding and resources in china. however, their performance is still out done by top caliber scientists who chose to reject participation offers. immigration retention h been a court u.s. strength with 91% of top chinese students with usaid doctorates still in the u.s. five years after graduating. long-term state rates are also high. amid ongoing strategic competition with china across dimensions including ai, the u.s. has several crucial opportunities to dance its own ai education and work force
pipeline. will close by providing a set of four recommendations for how the united states can boost its own ai competitiveness and these recommendations are first, starting at the secondary school level, u.s. education should evolve its focus from comput science education to additionally incorporate ai and an increase partnership with industry to bring expertise into the classroom. second, the national institute of standards and technology should work with u.s. industry to establish standards for ai and related certification similar to the process coordinated for cyber education. third, the office of establish hiring criteria for federal ai and ai adjacent jobs that are based on work portfolios and certification. fourth, congress should provide increased funding to the national science foundation, innovative technology
experiences for students a teachers also known as itest which shouldncrease the number of ai initiatives and develop k-12 students in relevant career paths. thank you again for the opportunity to testify before the commission on this important topiand i look forward to taking your questions. >> thank you. >> thank you. let me thank the commissioners for providing an opptunity to speak bere you. i wl talk today about the training education and deployment of semiconductor and i.t.alent in china. in 2021, the semicductor industry grew about 20% with revenues ring to abo $590 billion. with tt kind of growth, it will sn beco a $1 trillion industry andhat kind of growth will be anticipated in the future. it's safe to say the entire miconductor industry is faced withhe sam challenge, to
recruit and attract as well as develop and retain the qualified talent needed to sustain the industry in the years ahead. the arch for talent therefore is a significant issue among all parts of theemiconductor industry. the chise situation i particular difficult because it is playing a catch-up role as it trieto respo to the ineasing number of restrictions placed by the u.s. government a its allies. simply stated, without a sustainable pipeline of highaliber talent, china's gos for the semiconductor sector, espeally in terms of further indigeus asian will not be achievable. today, china accounts foa 35% share of the global semiconductor market, making it the e economy with the single largest share of that market. its efforts to develop more and more elements of the semiconductor value chain are dren, not only by itsnternal demand, bu also right nowhe
clear recognitionhat with the imposition of new stringen controls by the ited states, the country's only hope to meet its growing demand for advanced semiconductors must come from domestic srces. the situation regarding the supply and dand of talent for the semiconductor sectors is complex because of the suddenness which- with which the pressures ve increased. a region but likes have increased many of which are proving difficult to overcome. the answer is not simply be found in graduating more udents from chinese colleges and universities. currently, estimates range that the shortage todaynd into the futureill be somewhere between 200-300,000 skilled workers that will be needed inhis critical sector. 's also estimated rently that only about 15-20% of those who graduate with degrees in semiconductors actually work in
the industry. 2020, their team .7% of the 210,000 graduates chose to work in the indusy. there are many reasons for this. many graduates don't possess the righ sense- sets of skills that are demanded by industry. large mber of the schoo have faculty that don't have ample engagent wi the industry to understand the real needs of companies. a large number of institutions do not have pilot research or production lines for training students appropriately. since the big push for educating work graduates and specializing in semiconductor engineering and i.t. design began, there has been a big curriculum restructuring among many universities. in 2003, china established ic design and integrated systems as a major. in 2012, ibecame a show major.
in 2020, it became one of the 14 key disciplines that are high-priority areas in terms of overall educatn. this means the chinese government sees this as strategic areas for educational develoent. along with the status of ic -- i.t. engineering discipline, 28 colleges were designat to be created as national demonstration microelectronics colleges. a group of nine were designated in the first group and the other 70 will be given money later down the line. these are the best and brightest in terms of where china's student ought to be going. this training capability has continued to grow. in october 2020, the a drc and the administration held a joint conference about creating a national integrated circuit, industry, education platforms
designed to enhance the collaboration between universities and specific companies. according to a recent article one key problem continues to be that the industry and the demand for talent is growing so fast that companies cannot wait the 4-5 year cycle for graduate this has led to the establishment of a number of auxiliary institutions that are vocational and providing quick training so students can come online in the workforce very rapidly while the establishment of these traing schools is very important, we shod not loseight othe role of univerties. they are strategically important to the long-term solution. one problem is the difference in the curriculum between the u.s. and china. in the u.s., peoe going to work in semiconductors come from a broader array of disciplines including electrical engineering, chemical
engineering, and information engineering. in china, especially with the heightened focus on meeting specific industry needs, there is a greater emphasis on microelectronics as a whole with a little dose of physics. with some exceptions, death is phasized over breath. as more sectors become more complex, in the short tm it is usul butay not be ithe right direction in terms of long-tm need ere is litt doubt china has made a comprehensive commitment to address his talent shortage in terms othe entire value chain of the semiconductor and i.t. industry. achieving self-reliance is a strategic national priority. will china come up with a grand solution to me i tent nes for e semiconductor industry? china has ught to utilize the same solution south kore deployed to catch up with japan in the semiconductor sector.
it tries to hire retired or part-time engineers from socket. -- from south korea, japan or taiwan. recently, several simpson engineers were charged with violating samsung's ip regulations. taiwan stepped up its efforts to limit drc firms from poaching chinese employees from the islands semiconductor industry many americans working in china have had to leave the country. china's talent recruitment programs do not seem to operate comprehensive solution even thou there has been an increase in the numbers of people returned to china. the burden of following china semiconductor challenge holds squarely on the domestic education this them. we can see ample evidence of progress being made. it is likely china's talent in semiconductors rains a signicant problem f some
timen the ture area thank u. >> thank you. i amoing to turn to the commissioner. we will do this alphabeticly. vice chairman wong. >> thank you. mrs. peson, and lotsf your teimony is focused on e ai major ofocus on ai. ai is cross disciplinary, pulling fr not just cs but, i assume, since ssingata science, both engineering and eoretical disciplines. maybe i am under -- unaware of this,ut we n see specific proleration of an ai major to the sa extent as china. do y stillave an advantage because of the strength of the
education, particularly the higherducati system we have in these individual disciplines that me up t ai discipline? rong, you mentioned but strong eics and philosophies strong cognitives,his whole area of brain study that can be applied to the ai sphere. bacally, i am saying that in comparg us to china, aren't we stl at aadvantage? we do not have the sa push ma ai a major. it is stray from our western pepective to make this a discipline study. >> that is an excellent question. i will say first that there is no one agreed-upon definition or ai educaon. china has taken atab at standarding that and that is generally the approach the ministry of education takes two majors, to standardize them. fo thei major, it is
standardized under the engineeringategory. thos in the u.s. and china, espeally in china beforthe launch of the ai major, it was pretty similar in both counies in wch computer science departnts where the main breeding ground for ai challee. some of these computer sence majors in both untries have an ai concenttion or elective courses that can train you up. china, they have taken that a step further would not justhe ai major. they go beyond that to introduce natural language processing and computer aitions. to really get you into the machine learning anti-running techniques which are ait more of an optional direction in the u.s. computer science. i will note there are sever programs in the u.s.. there is no compreheive tally we know of that really multiplies how many majors are
but i will point to carnegie mellon who has a bachelor of science in artificial intelligence. the curriculum design also looks similar to what i have seen of the chinese approach where they are giving the introdtory method and hands-on tools you need to learn how to design and deploy ai systems. i would say thatwhile in the u.s., we are no intending a all for the u.s. to replicate the chinese approach as mentione in earlier panels, the decentralized proach is a strength of thu.s. stem an alws for morf the organic development and the discussion an integration of ethics. to the degree that is why my recommendations hone in on starting the entrae early. china has clearly had success in starti their ai education early and allowing students to alize it is a possibility to
work on those topics. to the extent that you can integrate th earlyn in the u.s. education system will help more people concentrate in computer science with a focus on ai or ai majors entirely. >> to standardize education and set up a higher education system in these majors and set of research institutes in the majors and public-private partnerships, is this focus on what getsore media attention on general ai or narrow ai or does the distinction not matter when you are talking about this level of research education? >> i have noticed they are doing both. there is ocean science and ai. those are more like application types. there are ones that actually focus a lot on ethics. but then there are also ones that are shifting into more future applications of ai like brain interfaces and things like that.
like wuhan university is a good exple of that. i would say that they are focusing on where applications being currently but also in the more forward leaving ones that we shoul be paying more attention to as they are on the true leading edge. >> tha you. >> commissione >> thank you. thank you to botof our witnesses. i nt to pursue what coissioner wan w going after and understand the qualy of theducation, as well as depth versus breadth. with things like self-healing chips, architecturon the core versus the edge and all of the varis thingyou know a lot
better than i am -- are the sations e u.s. has imposed and limitations reling tou and enabling technologies -- production technologies, what kind of affect is this having? do you think that the work around, if they can be developed, might create new technologies that we will be behind? mr. simon, do you wanto start? >> myense is that, while at thproduction lel, the impa of our policies has already been felt, in the research level, i think it has not had much impact the reason why is thiss part of the intertional science and chnology collaboration.
the chinese have as much access as we do to all the inteational journals d blications that are out there. they are also part of all the international scieific and technical bodies that deal with any kind of related issues. there is no dearth of infoation flowing in and out of china. there is a lot of work thahas been done to look at the top tier public educations in sciencand engineering that volve coauthorshiby chinese and american authors. increasingly, those are not ethnic partnerships. a chinese-american but also involve more widespread spectrums of americans and also different kinds of universities. i do not think the problem is on the research side right now. though, again, it will be
important to see if some of the latest develments and different kinds of chips and materials does makits way into mpanies. the fact lik they -- that they are pang suc close attention to indtry unersity partnerships and fostering an environment toacilitate collaboration suggests they recognizthe problem and they are onto it. it will take aong while for them to work. as iited in my larger paper there are numerous examples of this. the second point is, at the best universities, that is where this progss is being made. you have to understand ts morning that there is a real -- of the trains education system. somebody used a hundred universities and i think thais too much. if you go beyond the first 25-30, at least on the civilian universities, we drop-off is very sharp.
you can go to 100 of the top american universities in whatever field and get a pretty good educaon, a world-class education. in china, i thinkhe drop-off is sharp and reveals itself in the quality the fultynd the learning and instructional facilities that are available to the student. -- students. >> ms. peterson? >> i very much agree with what dr. simon id. that is why i highlighted th. even if it sounds shocking tha there is 450 universities even in 2 alonethe fact that -- in 22 alone, the fact that 100 universities have added this, theyre quite low tier univerties. it is interesting in whether they are just jumping on the bandwagon and do n have the resources or quality of faculty
to teach the conceptneeded. also on the employer's side and demand side, ty may not find the quality of education significant to employees that featured talent, even if there are talent shortages. they may not bite on that talent. there are a lot of anecdotal reportcoming out on the widespread issue that are being raised across china and the tech sectors the eort controls questio one thing that has not really been looked at tt deserves morettention is how this i gog to change the calculations of semiconductor talent, now that they have been severely restricted they are going to then choose to leav the country so that they would still beble to, in theory, access that know-how once they leave china's borders. that is something that i think would create a massive brain
drain and exacerbate the semiconductor challenge shorted even furtherdomestically in china. >> think the witnesses for your statement and contributing your expertise. no questions from at this time. >> i am going to pass and save mine until the end. commissioner halbert. >> thank you. ms. peterson, my first question is for you. have a number of questions i will submifor the record of my first one is one of your recommendations was to incorporate higherriteria and ai skills at opm. it tells with that ithat in the u.s. tech industry, talented ai researchers and engineers are often not working in the government becse they are offered millions of dollars to
work privately. it is hard for the government to compet with the private sector for height and engineering talent. elicit an example of the voice recognition, in-flight tech company that partnered wh several lleges to partner across china, can you give us an idea of what that looked like and the mechanism of how it worked? >> to answer your second question first, there are a number of tech companies that have created ai colleges. this is something that predated the 20 18th and i referred to that said we need at least 50 of these research institutes and colleges before 2020. i see the companies working most in the research institute side ofhings. they are really driving and taking the appcation scenarios that ty are aware of as the
mosthallenging things ac in the industry and directly bringing the problems in the clsroom. they help toesign that in mind and are also letting instructors and students ake in. these students can be dergraduates. the are usually more on the graduate se as iave se. they are -- structors are able to kind of work as companies like i fly tech and get t know-how a there is a cycle whe they bring it back into the classroom area students can go do training at these untrieand -- companies and bring the know-how back into the classroo have not seen formaliz pathways for studeso go work at places like f tech in the ture by would not be surprised if somef that talent then went to go work for thefull-time. this is something that ver much is not necessarily -- there is
the urban divide that playsnto this. whe there are maj tech companiesringing ai enabled education platforms to assist udent learning into rural areas, that is not the same thing as people working at ose compans directly coming to your classroom and teaching you about the cutting ge research findings and the indusy ndings. there, i see another divide in terms of the quality of ucatiothat people in moral areaserhaps cannot get because they do not have access to the real industry process. thank you so much. before i run out of time, a question i would like toevel for both of you is what's insight about china's efforts on ai or about how it is education promondustries do you believe toe true that mo people might disage with?
>> can you repeat the st part of your question? >> what insight about china's efrts on ai or about how it uses edution to promote its industry do you bieve to be true that most pple might disagree with yo so much has been written about china anartificial intelligence and semiconductors. i am curious to know if there are facts that she believed to be true that a lot of other pele may disagree with you o that dart fr the mnstrea of wt other people have already covered. >> from my understanding and what i see in semiconctors, to prepare for this testimony, i went and looked at what was going on state-of-the-art with respect to american semiconductor companies. i think the use of ai, machine learning tools, data scice and
all of that are now heavily embedded in the mainstream of our industry. as i read this, i started to look for do i see a lots of examples of this happening in china? i think china is at a very embryonic level in making this happen but i think the education system has strong recogtion that these tools are going to have to be brought into the mainstream education in order to make sure at graduates are able to meet the needsf this indury in e yearahead. the one thing we have to remember wch is ally crical iscience popularization in china something that is ve imrtant. because something is happening in ai or another advanced science area, the reality is the chinese government does see educating thgeneral population in science as something to be important for the long-term development of the civilization. that is something that we should
not lose sight of. it has -- thinking in a modern way, getting people out of this in the role areas where there is a lot of taboos and superstition, and to the modern world of the 21st century is something the government and ministry of education recogzes as the very important -- is very important to china's long-term velopment. not everyone growing up in a rural area is going to become an ai specialist but as a citizen, they need to understand what ai means to their lives and how it is going to be utilized. whether china is doing this in the best way possible, i cannot confirm. it is happening. it happened like that the chinese technologrevolution. as we came out of the cultural revolution, they were already 10 years behind it had a lots of catching up. i think this mass education
effort is something we should not discount in helping to further educate the population. probably not to the level we would think adequate but nonetheless, for china, it is working and having an impact. >> and q. -- thank you. >> want to actually follow that point but let me ask the question this way. i wonder how shoulwe think about the requirements of a certain national ai education program. because, what you descbed is really a massive cross effort from all levels in china. primary school, high school curriculum, ai majors. a skeptic might say this is what happens in a centralized system when the word has gone out tt ai is a big andmportant thing.
everybody at every level want to attach that label to whatever they are doing in order to get funds from the central government. you do wonder how many of these people coming out of these programs a actually capable of doing the kind of work that needs to be done to implement and make ai systems useful. i woer if this might not turn out in some waysimilar to computers where you have a relatively small number of can -- of people who design and a larger number people who build them a larger number of people who have to be trained to operate and matain the systems and then a vast population of people whose them but do not understand them or have to be educated to a high level in der to use them effectively. what would a right size ai education program in this country look like? >> that is an excellent
question. i share your system sentiments and i highlight the risks that they may run into challenges with shod implementation beyond the most elite universities. i would cite a more recent example like the whole chat gpt fevered. -- fever. you have all of these companies trying to support and develop similar copats even though they do not have any background in that field of ai whatsoever. like the food delivery giant is a good example of that. to more explicitly answer your question, i think there already are good programs in the u.s. and the high school level. one example that i would want pot you to is gwinnett county
georg. they have this interdisciplinary approach to ai that introduces some of the technica concepts. it gs studts interested in knowing if this is similar wh dr. son set of know-how hoai is affecting yourife but also get more of the rudimentary introductory chnical understanding. be able to devop that interest and potentially shape you working on those tops in the dergraduate lel and beyo. i would say, at the elementary school level, really introducing thiss how ai is affecting your life. this is something similar to initiatives that have been done for a long time from thesa trying to do outreh to help studts understand th is what cybersecurity is, and this is why cybersecurity is important.
this is how you can get involved and the different les you can do. similar initiatives for that, on the technical and non-technica side. emphasize the ai workforce is nojust the most chnically trained individuals but also those who are product magers and wo on the commercial side and the legal side. everyone having a better understanding how ai affect your life. there are certain individuals within that we have the most advanced typical training. i thinthis would produce the most robt ai talent pipeline in the u.s. >> thank you. >> both made efforts to suggest thats o2019 that was the last number with a very high percentage of top graduates in fields from china whwere studying in the u.s. and staying in the u.s. for at lease five years or lonr.
do you have any sense of what happened more recently since covid d more broadly, ab this is something to deal with in the rerd? what kinds of policies might the u.s. pursuit to encourage people to stay? just a quickuestion about the current data. dr. simon. >> the last number i heard the other day was 1400. that is the number of scholars post-doctorate, and senator read that over the last two years have decided to go back even though they had positions in the u.s. that is not a big number but not a small number either. a big question right now is many of these people literally are wrestling with whether they should stay, return to china cover or go to a third countr i have a number of examples of chinese friends in the u.s. that have been here, some for 20 years, who are talking about going to places like saudi
arabia and kuwait to new universities that have been built there because they can do their research and work. ey also can continue to collaborate with americans and chinese without running into political problems like they would if they stayed in one or the other country. i think we have to be cognizant that the recent events, with the violence in places like michigan state, some of the experiences of racism, suspicions under which there was under the chinese initiative etc. have really cast a dark cloud over the professional lives of many people. i think we underestimate the degree to which this has had a never give -- had a negative impact against these people. they are very cautious right now. even if we wanted to promote more sino u.s. cooperation in research technology, a lot of
these people would be hesitant. they do not want to communicate in some places with their former contact in china. i think we may -- i think we didn't go too far in termof the impact -- that we did go too far in terms of the impact we had on these people. i think we need to be careful. i mentioned the semiconductor industry is a worldwide problem and the only way we solve our problem here in the united states is making sure there's a steady flow of students from asia comin to the united states, willing to work on that. if we don't have that we won't be ae to akeve our goals. think about how many projects we have going on in the united states. where is that talent going to come from? how will we meet the needs of those companies being set up, new plants and research centers and all that? it demands a huge amount of
talent and i wonder has anyone sat town on the back of an envelope and calculad how our these will be met here domestically if we're not tang in chinese students andiving them h1b visas. >> i have several queson, i thinwe have some te, i hope we canet through them. . simon i'm interestedn your observation, your testiny, you didn't speak to it in yr oral testimony, that managerial requirements at factorsnd r&d levels have been gen lite tention in termsf education and we are very focused othe scienceide but when it comes toctuall training people w canng manythat there's a lack o interest. could you elaborate on what you see as the implications,
particularly for the semiconductor industry, particularly when it comes to training managers. >> i have been studying the electronics industry for four decades and i was part of the generation of people who were asked, can china innovate? no, china can't inknow vait. can cna produce state of the art? no, it can't. one conclion i came to is two arias of deficiency were one, the lk of professiol cadre of -- of r&d managers. in the chinese mility we have r&d managers who manage big defense projects and ey dersta how these worknd they knoeverything from budgetto personnel, et but in the civilian research side, that has be something that has been missing. we haven't seen that kind of ofessional cadre developed, nurtured, cultivated, etc.
you would have thought that it would have happened in places like the chinese academy of sciences and the 100 or so institutes under that, but for seven years, i actually brought a delegation of institute directors and deputy directors to the united states to train them in three or four-week buckets about the managing these kind of issues. and i would tell you, you know, there was a process called premature promotion in china. because of the impact of the cultural revution, a lot of young people had to be promoted into more senior posts before they were ready and therefore they did not havthat accumulated experience that one would have normally, let's say, in the u.s., by having gone through the ranks and reaching that level. so i think right now in the semiconductor industry, this is the big problem. if you looked a -- looked at it as a pyramid you'd say at the top are the most high-end,
talented, managerial personnel who run these things. but yohave to flip the pyramid on its head because the greatest need is also at that level. that's where they lack experienced people. i mentioned at the end they were hoping to get some of these people among americans and chinese americans, that were going to work in china from taiwan, maybe catch a couple of them in kor, etc. but that seems now to be -- that opportunity seems to be disappearing. so i tnk there's going to be a big gap in that area. i don't se anywhere in china on the technology side we see, i described it. i don't see anywhere in the management schools where they're actually explicitly prepari managers for high tech careers in industries lik semiconductor. the technologyanagement programs that exist in ina really, even though they try to co a lot of what's going oin the unit states, they really
don't catch what needs to happen. particully because the managers are not deployed on the ground in factories. they're too theoretical. they're not practical enough. soy worry that that's one of the achill heel in t chine efforts th they're trying to catch up. >> do you see that shtfall as a csequence of party dictate that they think poll toirks doctne or idea sol the apps to magement? what- whato you see athe reasons why people just aren't interested in magement? >> in 2023, the answer is no. i think they understand. there is no reason to believe that the operation of a smic plant or ymtc plant or any one of those is experiencing any
high-level interference. sure, i think that there are party committees in all of these organizations and whe xi jinping makes his major speech at the latest party congss, everyone probably has a half a day of study stetions -- segs to make sure they're cognizant of what the president has said. i think beyond that it's all business. i think that maybe at some lower level factories, maybe in more state-owned enterprises, we see some of that more prevalent. clearly not in the kind of state of the art plants wee talking about. i think it's all about business and i think it's all about the technology and competitiveness that they can achieve. >> we haven't spent a lot of time today talking about faculty. in both of your presentations, can you tala little bit about
where faculty are drawn from, where they're train, how they're traid, what the quality of that training may be, anthen wt kind oftandards do they take tts in ord to teach in iversities? at'she fulty like? >> so this is not an element that we have explicitly looked at as much so i defer to dr. simon. >> so let me give two perspectives. in the980's, 3% of high school raj was went to university. that numr then by t late 1990's grew 25% andoday, according tthe latest data, over 51% of high school graduates can go to university. you have to ask yourself, if the enrollments in universits are increasing so rapidly who is teaching these kids? that's really big question. and the reason y you have this
bifurcation is carly that at the top tr, they're getting not onlyheir own gduates, which ian inctuousness in the system. they tried to brk it down but thers not enough places to go. the second source of course i getting people to come back after theye been in the u.s. or europfor a mber of years. that's really why the talent programs have been so important for china is to recrt that faculty that are needed to staff these new positions across the entire system. and then also of course to dri the rernlg reseah. so there is big problem the faculty -- the st and the brightesstill stay in the ited states. and if th's the case, tn if you're getting t second tier, you know, coming from, i won mention any u.s. university but say it's not the ivy leagues that they're coming -- coming back from, etc. what ends up happeni is that they're notetting the kind o talent that they really want to
ha. that's what i they upped, durg the riod bore china's initiativeat leasting their had upped the stakes. they were trng to get people by hook or by crook. that's also why the program broke down. because they couldn't get people to make a decisionering stay or go. so they gave them a choice. come for a couple of month, we'll make believe you're here for nine months. we'll break t rules but we'll make it work so you can be here and you can be there and you don't have to give up your tenured b at a u.s. university. but the reality is they still aren't able to meet the needs of all the faculty. that's why the bottom-tier universities they're not going anywhere very, very quickly. but in the top ones it's just the opposite. that's why some made the top 30 or 35 of wor-class universities in the rankings. it's because their faculty are producing world class student and world class research. that is simply not occurring at other places i the country.
the a.i. major is actually not the only initiative that china has. there's actually 34 other majors that china has that have the term intelligent invasion in them. like intelligent transportation and things like that. it's more context-specific, depending on your fields of interest. the a.i. one is a much more general field of study. but that has only begun in 2019. so we haven't ev seen actually the first gup of four-year graduatecome out. they will comeut thispring and so it's similar with the a.i. institutes, those are both the research institutes and the colleges that are teaching the a.i. major and all these other 34 related majors. we haven't seen concrete
findings. it's a hard lynn to draw to say, this research was conducted that the institute and then translate it into the innovation game. but that's ongoing work i'm doing, studying what the research institutes are publishing and so are they focusing on more or less natural language process, ocean science, particularly curious if they are focusing more on surveillance or military applications. but i would also say that on a somewhateparate point we are also looking, measuring the job demand for a.i. positions in china and so we can see that there are, you know, the four major hubs that are classically associated with the top economic hubs in china. so including the beijing rion, those areas are huge vacuums for a.i. talent.
so that is also not necessarily to say that there is diffusion that is occurring but those are interesting areas to watch to see what emerges out of those hubs of a.i. activity. >> when you say vacuum, you mean concentration -- >> i mean they are vacuuming them across, from different regions in the country and leaving them to potentially have talent shortaglsewhe. >> i want to just add one point quickly. e of the proems that we se in semiconductor is this rapid turnoverf personnel. now if you're a company, you obviously don't like rapid turnover of personnel but if you're looking at diffusion of knowledge, the movement of people the high level of talent mobility, is something that could turn o to be beneficial. because people bring the -- legally or illegally or whatever, you know, let's assume for the moment they didn't steal
the i. but it's embeddedhat they brought with tm the knowledge that they have. they're able to upgrade wh's going on at the next place and the nextob place. that actuay turns out to have some beneficial effects even though the original company is not happy that they lost their job candidate or invested money in the training of their job candidate. there's this phenomena they say that there are hot companies and coldchools. wherehe schools are not intereed but the companies are ally interested. but there's theals opposite. cold companiesnd hot schools. schools that want to get their kids employed. but the companies are so nervous they're going to lose them after six monthsr a year because of the turnover rate. but the positive if you set that, it does provide a vehicle for difawtion of know-how across the economy. >> i think you anticipated my next question which is you mentioned in your testimony that 236,000 employees were recruited
from the market veus7,000 coming from campuses. and i noticed yesterday there was a story of the slide of several big tech pdd, ali baba, of their stock value because they are competing for employees. can you talk with a little more granularity about those two points? that it's just people moving within the market rather than coming out -- excuse me, coming out of sools into these companies. and where is there the most mobility? early stage career? what does it lk like? >> my sense is that the companies go to the markelace because they're looking for experienced talent. that's what they want. they don't like getting these newbies because many of them dot meethe initial exctations. i thought you cou do a, b, c,
and d, but you really can't do that. you heard this mornin some of it is apocrypha but they want to put you right to work very quickly to add value. if you can't do that you're pacecally t very rthwhile. i think they, the compa, the private companies in parcular, arin the marketplace. the state-owned eerprises, i think there's some more pressure on them to go to the universities and to hire those people. sily because they're much more part of the system. andherefore i think theyeel some obligation to do that. but even re importantly, they have -- they want t develop a steady popne -- pelineo they have people that are available even if they're not the best people that they've got people and they can work them through their syste i ink tt that's going to continue to be the case. i thinkhat we're going, as we see with the proliferation o companies as a whole, particularry on the i.c. design side, that's what's made that
market so lively, it's been such a big proliferation, they don't know where they can get the talent from. that's why these auxiliary vocational type schools have emerged. but again there's good news and bad news with those. some are very good. some are not so good. why? because if you're going to be very good you have to hire faculty who can teach at the state of the art. those people command a very high salary. -- sally. -- salary. which means you have to charge a high tuition. which meanses of you have to guarantee a job at the end or the applicant won't want to spend the money toet this specialized training. we had this in the united states with respect to programmers. particularly i know on the west coast but i saw it also on the east coast, a number of schools emerged, six month train, pay the equivalent of a university education for one year, but most of them, those graduates were working in google, amazon, wherever after they finished in.
china there's no guarantee like that. i think the students are dubious. the employersre dubious. so that's why in the marketplace, ty can have a wider range of people to take a look at it. so you've also seen a proliferation of head hunting firms who are basically looking for talent and peop tell me they get phone calls all the time. you'rehinking of ling --ot loseug moving your job, shifting your job. so i mean, it's -- the word chaos or chaotic is a good word. it is a little bit chaotic. it may be short term because the chinese now are feeling the impact of the u.s. restrictions immediately almost. it may be in two or three years it will sort itself out a little bit. supply and demand, you know. will seem to work itself out. i think thaall has to do with how fast they can improve the curriculum and the delivery of
that curriculum to the students. >>y questions ha been answered so leme ask everyone else if anyone else has any additional questions before we go back to robin. anyone else have anything else they want to ask? anyone remotely? ok, it's all yours. >> just one more question. since this is a hearing on educatn and training, and the futuref the workorce, i'm interested in you know, ms. person, w a.i. is being used to enhance edution and training. >>es. so that is -- that'a complemeary effort. i did mention in response to one of my quesons earlier that there areajor platforms such as baidu that is trying to bring a.i.-enabled education. it's not -- 's a bit of a
parallel concept where it's -- the virtual assistant type of approach. so that is actually quite widespread and bringing that to rural areas, that's meant to pose as an assistant, as a complementaryethod to ral teachers, torban teachers. and so measurg the efctiveness of that is pretty mulmost impossible. but it is somhing, it is also something that i would say is -- exists in the u.s. in a little bit of different form where that's a bit more in person. one of our research efforts has quantified the port foale of summer camps an after schools and things like that to enhance ample i. education as well. buthis is another facet that is not necessarily, i wouldn't say it's as explicitly called upon in the strategic plans but
it is one of the eects of how industry is trying to help build interest in sharpening the -- i think this kind of goes back to the first panel a little bit in talking about the lack of educational base that some students have especially in rural areas. so in theory, it seems like if it's the ample i.-enabled education methods are supposed to help fit each of the students a lile bit better and tailor it to their learning styles across a wide range of subjects, right, so like english, math, science, all of these other topics. i think that is something that might help them but it's still, at the end of the day, those really critical shortages and that very persistent ruralrban divide is not something i believe is necessarily solvable
through just access to a learning platform. >> under normal circumstances, the market wouldave been filled with a.i. summeramps and a.i. programs, etc. but after the crackdown on companies li new oriental, etc., there was a desirto make sure that there's not just, y kn, wealthy kids can go to these and the rest of the country is left behind. so i think w may see a different model develop where it's not just entrreneurial, you know, people who are tryin to open up a new business by creang these camps. i think it may become part of an auxiliary elemenas parof the education system. my sense is that the ministry of education has been told. do not forget rural education that we don want have a bifurcated country. we don want to have the 14
coastal city part of the country on the east coast and everything else. we know that in 1949 that produced a revolution. so preventing the bifurcation of the chinese economy in terms of quality of life, health care, educational access, etc., that's why going back to your session th morning, the big point about the role -- that should not be ignored. i was the executive vice chancellor for a university for five years, one of the four sino-u.s. joint venture universities, the one that developed our admissions policy for letting in the chinese students. i do have to say, even though there are some shenanigans in the backdrop, more or less the system has a high degree of integrity. fo americans doe we don't like it becau we can't craft a class. if you're at duke or you're somewhere like that,ou look
across, you look at all sorts of thgs in terms of the background of the students other than just the s.a.t. or a.c.t. test. some schools no longer even require those tests. but in china, the only way t get close to any kind of equitable access ishe use of the -- this kind of testing chanism. that'shy it hasn't disappeared. lots of stuff has been reformed in china and lots of stuff has gone away but the major vehicle by which you have upward mobility i china still is that, the question isow can they improvwhat it iss an education tool so it doest end up with students who have just rote memory. in the first batch of state of the unions we got at d.k.u., 60% were from china. that was the quota we wanted. the one thing the students told us, and d.k.u. is a liberal arts oriented joint venture university. they said the thing we fun i
the first several weeks is we re asked a question that we we never asked in o other education experience. i sa, oh, what's that question? what do you think? what do you think? so that becamehery, yo know. at do uhink a how do you tell somebody how do you think in an educated, informed way. that's something i think cna wants to cultivate, they just want to cultivate it within ceain guardrails but i think that the problem with china always is, it's never a 0-1. it's somewhere in the spectrum of finding a balance where we would be comfortable with the balance, they wouldn't be. but they want to find some balance. they clearly know that. and i think we see the improvement in the resrch environment, it really demonstrates the effort to give more degrees of freedom to the scntific and technical community to do their thing. that's why you have technological entrepreneurs, you
know, who have proliferated, even with the crackdown that went on a year, year and a half ago. the reality is that there's no dearth of people with creative ideas in cha. i think we would be making a big mistake if somehow we thought that creativity never emerged or is dead in china. i think that's not -- would not do us well as we look at the future of ourelatiohip. >> if there are no other questions, then in conclusion, i want to say, as i did at the beginning, we would like to remind you all that testimonies and transcripts from today's hearings will be posted on our website. i also commend to you the paper that was smitted by dr. hannam to the commission as well has interesting additional information. please also mark your caldars for the commission's upcoming
hearing on china's global interference on march 23. so i want to thank all our witnesses, both dr. simon, ms. peterson, for this panel and all our previous pams, the commissioners for all of your attentiveness and participation the staff who did so much important preparation, anwe are adjourned. thank you. >> thank y. [captions copyright national cable satellite corp. 2023] [captiong performed byhe national captioning institute, which is respoible for its caption content and acracy. visit ncicap.org]