this is two hours.

>> good morning and welcome to brookings. i'm john allen, and this morning spam is about economic political and security aspects of technology transfer. i want to welcome c-span to this panel. it will be covering us and our panel will be broadcast later today. and at the very conclusion of this panel we will be followed immediately by the next panel. there's no break in the process. just wanted to make sure you are aware of that. also wanted to announce that if you were unaware, if you haven't you did your cell phone, i would ask you to do that because sometime between ten and 11 washington, d.c. is going to test cell phone and remote device emergency broadcasting system, which means at some point i don't look down at the phone. >> that happens to be fickle in beatings, but let's did anticipate that and not have it be too much of an interruption.

i really have the honor this point as the president of brookings of hosting this first panel of three terrific panelists, dr. anthony vincy, richard, and nicole turner lead. anthony is probably the chief technology officer at the national geospatial intelligence agency, or nga. i never miss the opportunity to thank your agency for the terrific support that they provided to us in places like afghanistan and iraq and many other places so thank you very much for that. anthony has a long track record of success at the nga serving as associate director for capabilities and effect of plants and programs prior to his current role and essence of developing the agencies vital public-private partnership efforts and i think you'll talk a bit about that as we go on. a special assistant to the associate administrator for space technology of the national aeronautics and space

administration, or nasa. prior to his current role, rich served in a variety of top positions at nasa and most recently was nasa langley research centers office of strategic analysis communications and business development leader. as well as the chief technology at the center at the entire center. and nicole turner lee is a fellow here brookings innocent for technology innovation within the governance studies research program and a resurgent brookings focuses on public policy, designed to enable equitable access to technology across the united states. she's also an expert at the intersection of race, wealth and technology and comes to brookings after most recent survey as a vice president and chief research and policy officer of the multicultural media telecom and internet council and vice president and first director of the media and technology institute at the joint center for political and economic studies. we have three terrific panelists

this morning, ladies and gentlemen, and i'm honored to energy spent at to guide this discussion. we will be here for about an hour for the first 30 minutes i will offer some questions to the panelists and the second half hour we'll go out to you. i don't normally ask our paddles to do introductory marks, but, of course, in the first question that i asked which would go to all three of the panels if they choose to make it introductory remark they are most welcome to do so. with that let me go to the first question which is about technology transfer. starting first with the oval topic of discussion today, technology transfers, be they public good technologies emerging from u.s. government projects or university research, this is been an enormously, this has enormous potential ramifications when thinking of the growing tech race that we see as nations compete against each other and there is a big data and artificial intelligence

and in particular which comes to my as the united states and china, for example. let's go down the line of our panelists this morning and present the opportunity think it's an opening thoughts about doing a a new definition of public good technologies? and what's the u.s. government obligations here? is this important topic at a good way to begin this overall conversation. with that, anthony, would you like to offer some? >> yes. thank you for having me. thank you for having the pair which i think is extremely important at this particular moment in time for the country. i think looking at technology transfer, i put in the larger context of what is the appropriate role of government working with commercial industry and nonprofits and wider economic and civil society. and within that for my role at nga i think about national security and strategic kind of

consequences. over history and in particular since world war ii we have kind of gone to some phases of working differently with industry and what you with the public and we've invented new approaches to doing that. the entire idea text transfer as an approach, uses of contracts, approaches like an ota and things like this and i think right now we're watching kind of a geopolitical shift. and if you look at the national security strategy, national defense strategy we're saying the ship from, the post-9/11 world into a new era, near pure competition with china and russia. with that demand is a new approach to public-private partnerships and within that to text transfer. and seeing tech transfer as a

means of strategic competition, and within that i would include not just transferring technology but more broadly transferring, sharing, investing data and other intellectual property. i look at the last sort of 40 or 50 or 60 years of government activity, and i have noticed we built up a massive asset and resource really, whereas we used to think of natural resources is something maybe the government owned and been leased out say to the energy industry, now critical ip and data and technology kind of national asset. when you to come up with new ways to invest that asset and use it for strategic national security and economic purposes. when one of the things i've been working on at nga is to do that,

which is strange for intelligence agency to think that way, normally we are consumers of data and information, not providers to interview but i think that's what we're going to have to do to strategically compete and start looking at not necessary just something we open source out of the sort of public good. i think they're still a role for that and for open sourcing things. for example, the corona imagery was open-source for kind of historical and archival reasons but but i do think there is other aspects will be might want to not open-source it but still provide and may be treated more like proprietary information that we could unclassified of course we could provide to partners out there to universities, two companies to create technology. you brought up a eye. i think that's the major technology to source consider in this aspect where you need

historical data to train some of these algorithms. all of the sudden this asset that we've developed over the last 50 years of historical data is actually important in that economy. we have to find ways to share it strategically with certain companies not with everyone and i would suggest primarily with american companies or potentially allied. think about five eyes for example, if that's what i think we're developing at nga at the department of defense and intelligence committees and some other agencies also started work on. so again perfect timing on the topic matter and very important to what we all trying to do. >> thank you, anthony. let me turn to rich, from their position at nasa, , what are yor thoughts on this? >> one thing of what you just mentioned is that we need to make sure we think about tech transfer in both directions. as an agency that has a mission to do something technical, technologically pretty difficult, it's important for us to tap into the technological

activities that are going on outside of our own development. we have traditionally used tools like sbr i to tap into small compass to get their technologies to take advantage of those within the mission but recently we're also doing a lot more with prizes and challenges. with those we can reach into a much broader committee outside of just the u.s. and tap into technologist who made in a garage somewhere across the globe. this is really important as a look at technologies that now 70% of the research is done offshore, outset of use. for us us not to tap into that is a big mistake. we got to tap into that and the sketches of the partnerships that are to that kind of technology in order to accomplish the country thinks we are working on. i think we did make sure with a balanced discussion with the cookie tech transfer. the other side with regard to the transit subsidy something

that is in our original mission statement, something we do a lot of. something we have the spinoff magazine goes out the people to get all the time for the amount the thinks that come out of the space program. it's kind of part of our culture to do text transfer. i think the thing we're recognizing, , however, is this broad dissemination of the technologies is not as effective as perhaps doing it in a more focused way. we recently began some programs around working with individual companies come with individual organizations and particularly look at start of organizations to try to see how they can to prevent some of the technologies that haven't been developed within nasa, not only the technologies that take offense at some of the expertise come from the subject matter experts within nasa to help the companies move forward. we find we can leverage those technologies are a lot quicker and a lot faster and try to help that economic ecosystem be more

robust to actually transfer the technology in a way that is much more effective to businesses grow, to create jobs, et cetera. that's something we we're experiencing with everything is going to be something very important as we go into the future. i'll leave it there for now. >> that's a particularly important point about helping the startups to accelerate the process that might otherwise taken quite a long time. no call, please spirit take you, john. -- nicol. i feel so honored to be sitting next to these two scientistic uncle to talk about tech transfer from the perspective of civil society and what we are saying to a certain extent it is creating ethics of difficulties and challenges in tech transfer because the internet and the way it is been commercialized sort of excel with the private sectors engagement and tax the government sector when it comes to r&d. we all know the first tech trance was probably the internet and gps systems. and if we look at the way the system seven leverage as well as

u.s. regulatory decisions to make the internet more commercial, we are seeing which is a lot different today if we look back, where the internet growth is outpacing what governments can do. i think that's when referencing, how do you create different types of models of partnerships. i want to address this question what do we look at what was the public good technology, and just reflect on that for just a minute. this competition has created i think a good firewall where companies that have newly been created in the disruptive age are essentially, not mrs. o doing things for the public good. we had that first challenge where we're seeing the marketplace develop products in the commercial market that may not translate back into the public sector which is something that a record here at brookings, or vice versa we're seeing the government not able to keep pace with the private sector is doing

and i think that's somewhat problematic when we look at government funding to r&d, something that supported public technology. i also think public good technologies require some level of architecture that protects citizens. i think the second pair will talk more about that in terms of civil society but there's a challenge. it public good technology that aside for healthcare that's his eye for transportation, and by mental systems, military, do not have those protections in place which i think john, , good that your question, it creates a different definition of what we should be looking at look at public good technology. i also think it many of us in some have been watching the news, when you private sector companies that are edging into the public domain and suggesting many respects they're doing public good, there our problems and challenges associate with that. most recently with the breach i think all of us are very familiar with that is now toppled over 80 billion people. with respect to that i think

what nursing seeing intensive public technology, where we're seeing private companies like google, twitter, facebook, doing public purpose things, that didn't necessarily translate with federal agencies have terms of strict scrutiny around design, content, and the benefits of that product. i've always reminded of having worked with federal agencies, things like persistent medicine at some of the technologies we are seeing advanced to r&d, the question then becomes when you actually negotiate that are when universities, i've adapted to go to mit during the attack they were students are essentially putting patents for new products. it's a little different if you have seen the two screen television and think that's an interesting pattern on the product, out of university but when you start talking up drone technology or interference with national security systems or healthcare medicine to find itself in conflict with the private sector with public sector goals, i think it's

problematic. to your question i think we need to revisit a public technology definition, particularly in the u.s. as we see the framework which we started the internet from this commercial. our decision to make the internet commercial has implications on how that actually affect civil society. >> your thoughts are then that there has to be some semblance of transference? >> yes. >> and context of public good being used in the broader civil society? >> exactly. or the private sector will try that and we will have to catch up. >> terrific. anthony, back to you. in the world you lived in and currently do at nga, thinking about the american use every day, americans use everyday of gps and anthony is like google maps and what uber is probably doing in terms of using fake data collection individually, enhancing these technologies. but as we watch that unfold we

don't necessarily always know that it's for the public good. thinking less about the data itself and more about the technology that collects that danica should there be limits on u.s. companies that are allowed to sell this information abroad or even disseminate it to third or fourth parties on the context? >> yeah, i mean, this is a particularly important question. when i again going back to think about this as a strategic issue, international student issue for the country but i would say in terms of economic competitiveness outside of the national security realm, it's important. clearly there should be someone on technology transfer abroad. i think were all fairly comfortable with that and nuclear weapons, for example, come to mind as something that very clearly should be limited in how it goes. i think there is sort of the spectrum of what we're willing to transfer and should transfer, and some things do best when

they are fully open source and available to enjoy. so gps, for example, i think revolutionized not just one industry, multiple industries globally and definitely helped the united states economy, the government and everything we do essentially every day. you can't necessarily always predict when you look at a particular technology what the revocations are going to be when you do open it up. so my personal bias as somebody who actually came from commercial industry is that it is a bias towards opening up but at the same time putting on my national security had i realized that are some competitive advantages that want to keep within the country to support some of the industries here and support the national security community. and i think that is creeping

from, again, if we nuclear weapons all the way at the extreme, i think it is creeping left as more and more issues do become particularly important for national security and you brought up algorithms and date and hardware. i i think we're creeping towards the algorithm inside and were intellectual property should be protected. i think the primary issue to think there is as you kind of creeped left and you get into from hardware into software into algorithms, the shelflife, the half-life of these technologies become shorter and shorter. the shelflife of an algorithm might be months and sometimes even weeks. and so how'd how do you protect something like that, and is it worth, is the juice worth the squeezing you try to protected? had to kind of factor that in as well and a particular where much of that happens in open source and academic anyway.

those are the kinds of points where to start kind of considering. i would suggest what it means s taking a much more sophisticated approach to how we secure and i would think about securing intellectual property and technology in the country and thinking about multiple factors, and thin figure out where in that spectrum of transference we should allow it to exist. >> let me make a comment. i asked for your thoughts on this. the speed of government is woefully behind the advances in technology across the board, with its the production of datea or the collection of data processing of data. the emergence of algorithms, to your point about public-private partnerships, is there some hope that that concept of public-private partnerships can create a regulatory process that is faster than the speed of government to create the regulations that demonstrates a level of responsibility in the

private sector to do what we all hope, which is doing good and protecting the privacy of that sort of thing? >> that's an actual question, and something i'm intimately involved with at the agency. we do have a discrepancy, and asymmetry between the speed at which technology is developing commercial companies move and the speed at which the government develops technology or really importantly here, adopted and integrates into our operations, and how do we kind of shrink that asymmetry. i do think there will have to be some new regulatory policy statutory approaches to this. ask kind of what i i started, i think in our history we have adopted this new approaches when we determined that we need them. so into count is an example. we realized in the late 90s and early 2000 we we need to be able to communicate with silicon

valley and only started better, so that required new authorities. and more recently changes and how otas are able to be used in the department of defense. i do think were going to need some new approaches that kind of take the speed issue and i would call the speed of adoption. that really the invention because i think we actually reasonably good at r&d and admitting new technologies, commercial and she is great at inventing it and with great at buying it. it's how to adopt it faster. that's what i'm sourcing within the government now. look at things like project may even that's what they are focused on. i'm involved in project maven for that reason. and so i think that we right now can muddle through what i think it's incumbent upon congress to come up with new approaches that are going to support faster adoption. >> i think we will find, i

believe it's the 12, or the 11th testament on the help of the three tech giants. we'll see some of that down in the public domain what will excellent commentary. >> can i add to that? i think a complete we are going, we do see some of the public-private sector cooperation when it comes to big data analytics, et cetera but i do think we're in the state where much of the is a regulatory framework has been focus on consumer privacy, and i think the area we talk about privacy terms of text transfer has been more limited to the ip space at a think those conversations do need to happen. what's scary about this particular area and the rate of technology pace is it algorithmic peaceniks on the research i do it brookings is a algorithmic bias, and it's clear what happens in terms of the commercial sector, with data site is, there are not many agencies that of data site is on staff at understand algorithms and how to unpack that and are not many companies that want to give away the algorithm which is

what we talk about bias. because you can't really see what's under the hood, to go when it's disproportionally affected people. i think we've made progress. the gdp are last week with privacy week in d.c. we will see data protection and privacy policy would've come to the pie. i think the u.s. and about this blog after april 11 11th probay april 11 in the afternoon when he leaves the capitol hill testimony room we'll probably see privacy legislation begin to be debated. the question becomes with tech transfer, what we've seen in the last few months is the manipulation of what is available to sort of innovate new types of practices and procedures. with the algorithm has manipulate democratic institutions. that's a different type of regulation with the data flows are not easily identified, and not necessarily understood by all actors. i would even say in some cases

the private sector, outside of the silicon valley, i would say companies who were expressing these breaches and we don't even understand what that means. i believe going forward we will have to look at this application that distinguishes between consumer control or consumer access to their own data, business or enterprise access to data and then the government transfer of data as three different verticals, that some would have to be reconciled to create a much safer and resilient system. >> let me just add. maybe it's a bit of bouncing and we talked in the pre-conversation we had can we talk about the balance issue between these issues. there may be replication some of these new technologies but also opportunities of these new technologies. we've got to be careful not to over restrict to miss the opportunity. that's the balance we have to find the right place in the middle.

as listed this idea public-private partnership is a really good one and we've done that with the airline industry before to try to move composite out into the industry. that serves, they have the speed that we don't have an order to move those technology forward and take advantage of that speed is very, very important. nasa is a pure oxygen is like the rest of the government. as china move things forward is difficult but when we can have the public-private partnerships they bring the speed in, they make it happen a lot quicker and that's important for us to move it forward and within the opportunity space. that's important for the future. >> rach, thanks for that contribution. let me shift over to you with respect to your nasa background. we've seen the growing success of companies like spacex combined with a resurgence of interest of human travel into outer space and in particular towards mars. what are your thoughts on the private sectors will on that? are the risk associated with the

technologies are visiting developed in the private sector in the context of technology transfer? >> let me make sure we got the right perspective on this opportunity, frankly, with regard to organizations like spacex, blue origin, et cetera. let's take us back a few years nasa was formed out of an organization called the aeronautics committee. that organization was really put in a position to try to help the fledgling aeronautics industry move forward, right? they had a lot of research. they had a lot of policy discussions about how do we open up the airspace to these crazy companies that are flying airplanes around and oh my gosh, do we have to have some kind of restrictions on those, et cetera? that is the job that nasa in its history has been a part of for all of its lifetime really, is trying to help industry growth such as that it can commercialize and become something very valuable for a

country and, frankly, for the world. and so we're in a new era of that now with regard to this space economy where we're seeing organizations like spacex, blue origin, et cetera, et cetera,, right, for taking similar technological developments that nasa is that over the years. taking them on a together how to do them cheaply. again this is that something nasa is good if you're we're good the technological stuff. we're not good at figuring out how to do it cheaply. spacex and is configured how to do it cheaply sedgwick and didn't have a commercial market available. this is positive from our point we think this is, this is what should go and we are very much working to support those industries. they're using a lot of our facilities and capabilities, subject matter experts. we're actually trying to become a customer of theirs to try to get them privations up to the space station and eventually humans into space. we see this as a very positive benefit. it is the public-private

partnership with the talked about in kind of a big mega economic way. >> if i can come back actually can you bring up a great point, which is relative advantage and relative competitive advantage and what is private industry good at and what is the government good at. that gets to the heart of why you would even have public-private partnerships. when i look at the government, in particular to get agencies like nasa or nga or the department of defense, it's very good at doing certain sort of even more or less possible task, putting people on the moon. i think of the corona program, putting satellites at the american literally dropping fuel canisters over the ocean city can apply to pick it up and then in mid air and then getting it back to the u.s. to be processed and analyzed. that's just newly merely an ime feat and its incredible they did it with the technology they had. where as when you start to look

at private industry is very good at something very, very different. the inexpensive is a big part of it but also being created in a way the government isn't. an example i like to use his ways. elected government approach, for example, to global traffic monitoring if we had to come up with it ten or 15 years ago would've been to go out and buy helicopters. we look at how do you monitor traffic? the local news stations can make use helicopters and a film the traffic and the radio it down. so we should spy -- we should buy buy a lot of helicopters and you ellipses around every city in the nation and radio it down and let's go global with that. whereas ways with almost no money, and without even that in mind creator a commute of people to kind of go for our mapping purposes to start within for monitoring traffic ensuring that. and they use cell phones and use

gps and now they give it away for free. that is just a very, very different approach for problem-solving that lynn's itself to very come to solving very, very different problems. and so when we start to think about these public-private partnerships we should look at it from that lands of what is the relative competitive advantage of each side and what should they play? even within that example we see an example of that is gps, right, waves could not of been possible without gps gps cannot even possible from a commercial perspective. it's a money loser i suspect and it's really big and expensive and complicated, and just difficult to do and it has to last forever come for decades and decades which not all companies do. that's a clear role for government but then there's a clear role with something like waves for creative new uses for it. so again it's that relative

advantage and it's a new way to think about how government partners with private industry. >> and just i don't sound like debbie downer, i do agree the panelists, positive ways to deploy technote and i agree, spacex is nothing looked at to provide broadband services to rural communities. so there's this report is a technology that is going on ever deeper i just think this panel is responsible not necessary standards. but some type of input output with the commercial and public sector partnership is defined of what the output is so that it doesn't become something where you do have geolocation in the light of the government kicking themselves because the geolocation has an unintended impact or consequence that was not thought of by the commercial sector. i wanted to make sure, everybody knows that i do support it. >> it balances. >> it is. and i think that, threat and will go to the audience in just a moment, the common thread is the public-private partnership is really the way ahead here.

it is something that can reinforce the public good. it can probably minimize the deleterious or negative effects. waves has the capacity as anthony set of producing enormous amounts of useful data with respect to metropolitan planning infrastructure planning and that sort of thing. but it also will tell members of my family and i stopped off on the way home and i were slightly about that. with respect to commercial involvement in the space program, i absolutely agree with rich. this is been an accelerated. it's also very cost-effective. it's a quick integrated of technologies. the one tesla i had my eye on is on its way to morsel out of the business for a while. but i think this is real opportunity for us in terms of technology transfer both into context of national security but even more so towards integrally towards the good of civil society and the transference to civil society that is the real

opportunity. you touched on that and a number of times in terms of opportunity. any go to the audience. we have a rich array of attendees this morning from a number of different organizations. and for many different countries so we welcome you here today. we will go for about a half hour. i will and straight on the hour at 11 so i 11 so i apologize fg up five minutes of your time. if you could give us your name please, where you are from, and if you get to a question relatively quickly i would be most grateful. if not, i will find a question in what you were saying. so please, yes, sir. third row back in the moon, to the second row after that. >> john, thank you for very good presentation. my name is elliott, i'm from rockville, maryland. i used to work for the world bank. good people please define the

public-private partnership a little bit more clearly? >> why don't we start with you? >> it's a loose term and it's been used in a lot of different ways, historically, everything from building highways to nih investments in health care and so forth. but the way i would use it is to see it as, what are mutually beneficial things that the government, the public side, and the private side, primarily commercial industry but i would also include universities, nonprofits within that come what are mutually beneficial things that they can do with specific projects? and those benefits might be very

different from one side and the other. so, or they might be the same. so the example that i would use is the co-creation of technology what i was saying before, the government is good creating certain technologies, the commercial sector is good at creating different kinds of technologies. how can they work together in a mutually beneficial way to come up with technology that they both want? and so, therefore, shipping and different things at different times. and i see it as a partnership in the sense that it's not a one-way street. so it's separate from contracting, for example, where the government provides money and in return gets service or products of actually more of a street. ..

anyone else? >> i want to add that i absolutely agree. our public private is often in that vein. the other kind is we do a lot is this spring competitive work where we have several companies come together we will partner with as a team to work on technology and the maturation of technologies to a certain level that is be competitive and the these can take off in do with it competitively so both of those are models that were used. >> and i would say put with public-private partnership is an area that i do a lot with in the telecom space is you have to

have common goals between the private sector and the public sector in what you want to accomplish and a fundamental interest in protecting the public interest, honestly. public-private partnership is not a partnership but a public interest or civil societies but it's not at the core of what that partnership looks like and i would agree with rich on this that the public private partnership has to be done in a way where it does not stifle innovation. you seen the arrangements where you go into the public sector private sector partnership and concerns on whether or not you can have the ideation process happening in innovation can happen because of the constraints of either the public sector or the public and private sector in unwillingness -- electronic cupboards are a good example. we seen electronic health records become much more resilient because the public interest guided by hipaa and other sort of regulatory prescribed rules have helped private-sector innovate in a way that more readily available and i would end by saying it's the

scalability of that partnership that has to generate the output and we often deal with at in my particular workwear were asked to see is the benefit to civil society benefiting one part or a block if it's not scaling in your taking the money for competitive advantage is hasn't met the criteria. >> thank you for that question, mr. horwitz. question in the second row, please. >> the left side of the room seems extraordinarily inquisitive this morning. [laughter] we invite anyone to take questions from the right side of the room. >> thank you very much. i'm an african-american journalist and this is exciting to be at the dawn of the digital era and watch the differences that will bring for our societies. could you please make put into context how this relates to future government up until this point really the societies

worked on a global system of opening economies and coordinating their economies in governments. right now we are in the midst of such populist movements and with technology being at its earlier stages although we are seeing some of its fruit we are also seeing being used and we have described this process of where it will take time to produce relations and perfect techniques to ensure that they still serve the public good and government can do their jobs. can you put it in context. this will take a long time and yet we are on the brink of what looks like a nuclear war and so many conflicts brewing both in the public sector and global. thank you. >> who would like to take a crack at that? >> i think your question is spot

on. i described this as a myopic tendencies of creators who want to see a product to market quickly. a lot of the tension we discussed and the broader goals of what that technology's impact is on society and you are correct that many more global and i would argue that transfer can become protected by people in terms of what they want to share because the vulnerability that it created by having the technology so much more widespread is not generating the outcomes that we thought it would generate. different from a government for leveraging technology and practices to be more efficient to cutting into your democratic institutions i really do think will go into an age where those regulations happen before the innovation catches up which will be a flip-flop model of what we've seen. i think government will come in

and put a plug in some of the stuff before the innovation comes out which might also create its own set of problems. i think generally your question is correct. we have to be your ways to harmonize the systems so that since we are moving into more of a digital economy the bureau of statistics put out that the percentage of gdp governed by general is certainly something that is taken up a huge portion of our attention not only in terms of the innovation but economically so this will be a problem and it's a good problem to have but it will have consequences. >> let me add a couple points. there is enormous capacity for the community of nations using the digital environment to accelerate civil society, economic productivity and even improve government in many ways part from the come security side

of it. here's where i think the community of entities which is bigger than the community of nations when we talk about entities i talk about facebook, google, some of the most significant sovereign entities when you think about sovereignty have the capacity to help the slingshot in many countries that are in the developing world that would not otherwise be able to do it on their own through public-private international partnerships have the capacity to be quite helpful in that regard. you mentioned the potentially emerging cold war part from the dustup we're having right now the chinese there are enormous opportunities again for us to share in to cooperate the us and china on a number of issues. the issue with respect to russia and i have great concerns about where all of that is heading, brinkley. i see china in a different mode and i think we beyond the reflex

for protectionism which can chill this and can chill the opportunity in other areas the capacity for the united states and china in this regard to find common ground in the digital future i think is extraordinarily important to pull countries into the community. i would hope that we don't confuse the activities right now that appear to be protectionist and that could lead us down the road towards a trade road as being helpful over the long term for the us china relationship. >> if i could add in there, i completely agree with your point in the points that come up on the values and i'm a technology optimist. i used to work for alvin koestler who wrote quite a bit about the subject and talked about how you can skip a generation of technology and have profound effects on countries to be skipped from

landline's two cell phones much more resilient and much less expensive technology that i've done a lot of my phd fieldwork was in africa and i watched this happen in real time what mobile phones did to an entire continent and that really is the promise and we all know there are downsides and there's a negative consequence for a lot of technology but the net gain to me is massive and i think we talk about the public-private partnerships there's an onset assumption which is really important which i see in my government role more and more which is an acceptance by government at the working level that commercial industry, commercial technology is a lot to contribute to governance and government in general and that we should adopt it and this is a very real factor within government within any large

organization. i think we're starting to get over that and realizing that technologies that we can use that allow us to govern and to enact our government duties and responsibilities significantly better, cheaper with a wider scale. what came up before, for example, broadband where government again in the ways and maybe we don't by helicopters but maybe we do to go hole in the ground to every house in the country the matter how far off and that is valid and everyone in the nation should have broadband but all of a sudden there's a game changing technology that if you do broadband from space because of a company like space x and other ones that have come up with a new, much cheaper, solution and that is the promise here and we need to be open-minded as government employees and thinking about that. >> we are not likely very

quickly to these questions. the gentleman that is next to the camera, please. yes, sir bobby from no one. i would be interested in your thoughts about the wealth that has been created through digital technology. that wealth is unequally distributed and the digital industry is becoming increasingly sophisticated at the creation of roles as other industries have in the past once what has been created. do you think we have the right balance in the public, private partnership with the financial wealth that has returned to government for public purposes that public governments can make decisions about? >> no. [laughter] leave it at that. no, i think, i completely agree

with you. if we go historically and we look at the evolution of technology in general and something that i always have to remind myself we started with it in very basic principles of what tech events. even with government going back. that has morphed into an economy that is not just been a static economy but internet is no longer this websites where people go to but layers that create in and of itself its own wealth so the sharing economy outside of the general economy which is ruled by the internet and other really cool technology that you can touch, cloud computing, each of those layers generate its own sense of output, economic output, that does create this unequal

distribution of wealth and access. all of this in this room we don't own a patent or technology companies are what we all should be proud to say this but we are all passive consumers in this digital economy right now. startups have incubators and government is trying to break through and this is a case where government is behind on that, too. the season were government driven and or pcs that were supported public technology applications are now just coming around to see that we've got to fill this role mess roerole impeding its own generation of wealth in its own generation of workers which is why we are wrapped into this conversation which can be so much more sophisticated. who are the people on the end that will become the colonies of this digital economy? the 11% of americans that are not online, the poor, the

disabled, the rural communities, they will always be involved because their big data is what drives the new economy. even if they are not online. their lack or absence of data drives company to know where they need to deliver food what kind of investment that the smart grid system. they are still part of it but they run the risk of becoming deeper and deeper in poverty and eventually becoming digitally invisible. to your point i think the -- the bls was the latest case to debunk what the digital coming look like but your question is critical particularly when government invests in r&d and they don't get a return back on investment or technology which was designed to solve the social problem actually creates a problem.

that is where i think again a lot of us are stuck in the middle and the next time we'll talk about security but how do you begin to resolve and reconcile and create harmonious legislation or regulation that allows one part of the technology sector focus on civil society where another part of it continues to do what they do but perhaps in regular the context because again, the communications act of 1934 was designed for the telegraph later picked up by broadband and didn't anticipate the companies we see today. i think your question is a spot on and 12 equity index will continue to widen based on where you are within the topology of a digital economy. >> that is a great question. let me offer a couple advice. the sovereignty throughout much of modern history has been shaped by the concept of westphalian society hundred and 70 which the line on the ground that circumscribes terrain and certain number of people often with a homogeneous identification all who provide their loyalty ultimately to the

sovereign. that's a modern relatively modern view of the concept the west alien concept of sovereignty. somebody really is about the capacity of a sovereign to influence and that's the traditional sense all the way back to aristotle. i think we need to think differently in the public private partnership concept may be some of the thinking of how in a world where tech giants digital giants control the modern version of the power of ancient sovereigns which is wealth and data and algorithms in ways that traditional was feeling governments don't necessarily control them and we need to think of it differently now about public private partnerships ultimately morphing into what might be public private alliances because when we think about the larger tech giants with gps, if you will

that surpass many of the countries on the planet that can reach out and touch people in the numbers of billions and influence their thinking for voting purposes et cetera that's a whole different way of thinking about influence and sovereignty in the modern digital era because where it was in the past about terrain and numbers of population today is much more about the information that you control the way you will wield that information in terms of algorithm and the outcomes that will flow from that which will be prosperity and wealth. it's a different way of thinking and i think we need to expand our view about the use the term a moment ago the community of entities we often talk about the committee of nations and it's much bigger now than the community of nations and its community of entities joined forces in a public-private partnership and we have real capacity i'm not sure that were thinking about it properly and i'll leave it at that. >> do you mind if i build upon

that because i think what you have done is place this concept of public-private partnerships within the wider spectrum of how we understand geopolitics and how we understand history and i think we can go further with that thinking in terms of how different nations have different government systems and that difference will apply to how they use public-private partnerships in the nature of those relationships so i think china will have a very and already does have a very different approach to public-private partnerships when the government deals with companies like [inaudible] than we do and how our government and the us government deals with a company like google. i think both of those types of relationships will shift over time and that is what we are starting to see now and what were talking about here but what makes it even more complicated is these companies as has been

brought up here aren't necessarily just within a single government system and so google may have a very different relationship with the eu that may have with the us and it may have with china and it makes a very complicated to think about how all of these relationships will interact and evolve over time and that will be the sophistication for us on the government side how what do we actually want from these public-private partnerships with these other entities and how we will get it and that is what we are starting to think about right now. >> this is an important outcome for this panel and importantly my cable advice tells me we have four minutes left in this panel and i've been very disappointed in the right side of room to this point. [laughter] -- is there anyone -- yes sir. i like to get quickly to a question invented the 20th gentleman in the third grade. >> thank you so much. i'm from china daily. in terms of cooperation between

china and the us could you elaborate more areas physically where china can improve their cooperation to the well-being of countries and you talk about the areas where the government on the one hand to set the regulations other different industries but in the meantime need to provide sufficient funds for different industries to develop and how does government to strike a balance in terms of the us? thank you. >> that's a very long question with a very short question but a long answer. there comes to mind immediately where the us and china in the context of the digital environment can cooperate would be in medical diagnoses and the capacity to harvest enormous amounts of information on medical research and using the right algorithms to help us get

more quickly to diagnoses now that were are beginning to see can be harvested and rendered with high levels of competence. the whole business of countering terror and there are other reasons why china and the united states need to cooperate in this regard and there are real capacities in that regard, as well. i would sibley say we have not seen it play out yet but we are very interested in the same how president xi's objective with respect to the outcomes of the 19 party congress and china's intend to surpass the united states by 1930 with emergent knows how that will play out. china has some cases advantages some people would say disadvantages in that it has at its core the capacity to create great cohesion between the objectives of governance

government and the objectives of chinese companies. there is much more capacity there than perhaps in our system and i don't call that a strength or a weakness but different in what we need to acknowledge is that it will be different for us. anyone else? >> in the area of ai clearly the artificial intelligence side of what we're seeing and it's interesting and outpacing some of the things that we have here in the us but when it comes to global decision-making and problem-solving i think some of the applications there should weren't cooperation because particularly when the us takes on issues like a couple years ago in the white house trying to solve it stop the ebola crisis some of the new applications of emerging technologies could have been more helpful if there have been more cooperation and we are starting to see the human do more of that kind of operation as an international when it comes to digital rights and so bright they should be moving into this conversation around but i would also say to your second question around on that and the interesting conversation we had tabled is to incentivize

and regulate but you incentivize these digital development. hopefully we will begin to see more models like that were there are cases where we can incentivize government to have more where are there cases where we can incentivize governments to have more cooperation around products and services that fit sort of the core verticals of the public interest, which i think will help with the allocations of funds. because unfortunately, i don't think any government has enough money to support the local gdp of a company that's, you know, surpassed the gdp of a small country. so i think having more of that. i also understand for example in africa, there were lessons there in terms of the wireless boost. so you've got to incentivize those types of experimentation or projects to sort of balance the regulatory framework. >> nicole, rich and anthony, thank you very much. ladies and gentlemen, thank you

for joining us this morning at brookings, and would you help me to thank the panelists. [ applause ] [inaudible conversations] >> while mike does housekeeping appear is okay. while mike does some housekeeping up here, he's a very generous host in providing water for us, i will introduce our topic as well as the panelists. so i'm daryl west, vice president of government studies here at brookings and also the director of our center for technology innovation.

so we are going to continue the discussion started by john and his panelists on technology transfer. i do have to say the first panel set a very high bar in terms of both substance and humor. i'm not sure we're going to be as competitive for the latter of that. and also worked in references to aristotle. we may or may not meet that threshold, as well. but with he will try and get into some equally important issues. i do want to remind the audience both our c-span audience as well as the people in our auditorium, we have set up a twitter hash tag, #techtransfer so if you want to make any comments or pose any questions. #techtransfer. so our panel will focus on the security angle regarding technology transfer. we'll be getting into questions such as when should technology be transferred, and when do

sensitive products need to be protected. and where should we draw the line between national security issues versus free trade and the free exchange of information. to help us understand these issues, we are joined by a set of distinguished experts. to my immediate right is heather roth. she works on the ethical aspects of artificial intelligence. she has published several articles on autonomous weapons and is author of a book entitled "global justice: kant and the responsibility to protect." so if they had air stat he wiai stotel, we have kant. and we have the sydney stein jr. chair, author of numerous books and works on defense strategy in american national security policy and serves as director of research.

paul tree joldo served in severl senior policy positions within the u.s. government over more than 25 years. chris meserole is a fellow in the center for middle east policy at brookings, an expert on religious and sectarian conflict and impact of technology on foreign policy. he also is using machine learning to study violent extremism. so why don't i start with heather. you have argued that many emerging technologies are what you call dual use in nature. meaning they can be used both for good or bad purposes. how should we think about technology transfer and expert control with dual use technologies? >> thank you, daryl. thank you for having me here

today. so i think to answer your question, there's two ways to think about this. one is to talk about kant. one is to talk about hobbs. if we're going to throw down, going to double down on some philosophers. the primary purpose of the state, job one of the state is to secure the rights and lives and protections of its citizens. this is very hobbsian, right? the whole job is to actually protect the body politic. and so when we want to start talking about technologies and the regulation, what we're really talking about is the regulation between civil society, civil applications of a technology for peaceful purposes and serving the economy and something used for military purposes. the flip side of that is that same technology could be militarized in a way and used for security purposes or for weaponization. so we have to be very careful about how we draw lines around these technologies. one way to think about this is through a series of arrangements we already have in place. we have international treaties like the missile control regime.

so the mctr. we also have things like the vassnar arrangement, a voluntary arrangement of kind of like-minded states that want to ensure that the technology developed and exported is -- when a dual use technology actually starts to become so precise in what it can do, it becomes more conducive to military applications. so thinking about things like hardened systems against electronic magnetic pulse, so those systems that can withstand a nuclear attack or those that might withstand extreme temperatures. when these things start to happen, those begin to become what we would consider dual use in need of export control. so the good and bad purpose is one way to kind of think about the hook. but the other side to think about this is military versus civil.

and then it depends on which side you think is good or bad, right? what i would say about technology -- about tech transfer and expert control is that the new emerging technologies are not very amenable to the current structures that we have for expert control in dual use. so if you look at something like the vassnar arrangement, for example, you have within about 190 pages all sorts of discussions about what needs to be regulated, at what rate, and if it's this type of technology, if it's nuclear, software, enabler, if it's a sensor, if it's a frequency hopper, all sorts of types of technologies enumerated throughout that 198-page document. within three lines of that document, i found something very anomalous and i think is very interesting. and that is voice encoding. so if you can voice encode, so you can take continuous speech and then you can make it into, you know, 0s and 1s and encode

it into a digital frame and you can compress it and compress it and transfer it at a very slow rate. 700 kbs, kilobytes per second. that's a slow transmission rate. and for some reason vassnar finds this a dual use good that needs to be regulated for export control. i don't know when they decided this was the case. and i don't know exactly why. but what i do know is that about six months ago an academic decided that he figured out how to do voice encoding at 700 kbs and dropped an open source on the internet. so that kind of move really pushes us to think about how we do our regulations. and how we can be more -- have more foresight about our regulations when it comes to military applications for security purposes. another thing that i would just really briefly draw attention to is not only do we need new governance structures and new ways of thinking about types of security and military technologies or technologies

that, hey, i want to regulate this algorithm, by the way, that's the same algorithm on your phone running siri, and be then you're going to stifle innovation, and you're going to stifle the ability of other states, particularly developing countries, to gain that technology, to boost up their civil society's well-being, as well as their economies. you have to be careful about where you draw those lines. but there's another question, actually, that's equally at play. so you have, you know, where do we draw the lines, how do we do new normative governance and forward thinking on really hard questions about these dual use goods. but another one comes when these public/private partnerships happen in the security realm. and i think one of the things that we can look at now, it's been in the news quite recently, is the potential for a cloud computing contract for a single company to take all of the dod's data and host it on the cloud for the next ten years. single company contract, ten-year span. that's a lot of money.

and right now in the news what we've been seeing is amazon is up for that contract. we don't really know if they're going to get it. but there's been lots of discussion about amazon cloud and amazon's web services hosting that. and then we have questions about that public/private partnership and what that does from the civilian side of things and security side of things when a public global company like amazon starts making bets that it's going to host a defensive -- a state's military data. right? and what that does to, like, where amazon operates in other countries that they say maybe i don't want to -- i don't want amazon to have my data if they're going to be feeding it to the u.s. government. or those types of things. how are we going to partition that. how are we going to keep that export. how are we going to keep that dual use goods. make sure the public/private partnership is for the good of everybody if the states or entities are global in nature. i think we have some really hard questions when we start thinking about technology, civil military, and the securitization of all of this when they're

running together in really difficult and complicated ways. thank you. >> okay. thank you, heather. so mike, you work on great power competition, particularly with regard to russia and china. what are your thoughts on when sensitive products need to be protected? >> thanks, daryl, and good morning, everyone and i certainly won't try to rival john allen, our boss, on humor or anything else, but maybe will try to rival him on zigzag pricing from a brookings podium. much of what president trump is trying to do with china is justifiable and not necessarily in every detail but the general thrust of pushing back on china in particular. because let me try to have an historical perspective, not quite far back as aristotle, but if we think about the last 500 years and paul kennedy wrote about this. we have seen european powers in particular rise and fall fairly fast. partly because they couldn't protect their advantages, because they were living in an economy that didn't have

barriers to technology, transfer or theft of intellectual property, which is an age-old phenomenon we have seen long before the internet. and in just the last 50 years or suh years, britain lost its advantage to japan, germany and the united states, among others. but those three in particular. luckily, one of the three was us, and we ultimately therefore were in a position to help bail out the rest of the western world in world war one and world war ii, which resulted partly because of this technology transfer happening fast and germany in particular catching up faster than it might have otherwise. so i put all this in perspective, because, of course, our more recent historical reference point is the cold war. and we had very little economic interaction with the soviet union during the cold war, and we were very comfortable putting up a lot of the barriers, some of which heather just talked about, some of which were done

to prevent nuclear weapons from getting to other countries. but many of which were designed to keep high technology, conventional weaponry out of soviet and war-packed hands. now we're living in a world in which our most likely competitor is china, which is, of course, so fully integrated into the world economy. and we made a gamble not just in economics, but strategic terms, 10 to 20 to 30 years ago. we decided to try to bring china into the western economic world as fast as we could, including membership in the world trade organization. and the gamble was on both economic and security fronts, this would help china liberalize fast enough that the risks of seeing china grow fast would be outweighed by the liberalization of china, and that it would become a more rules-oriented participant in the international order. this sounds like china-bashing. i'm really not a china-bash er. jim steinberg and i wrote a book about how we could get along better with china and do things on the u.s. side that would

promote that process and recognizing china's impressive historical rise. but in some ways, it's become a little too fast for comfort, and especially because of the meanings that china has been using. daryl, you and i talked about the idea for this panel. i want to thank you for the whole concept for this event. we talked about this a few months ago when we met with michael brown from the diux unit in silicon valley. and they wrote a paper which is very compelling called "china's technology transfer strategy. how chinese investments and emerging technology enable a strategic competitor to access the crown jewels of american technology." they went so far as to suggest to rethink how many visas we give to chinese students studying in the united states. i'm not sure i support all of them. but their analysis is pretty compelling. it's recommended reading for all of you. and i'm just going to begin to wrap-up here by saying some of the ideas put forth to try to force china to comply with the

rules-based order for technology transfer i think are actually appropriate, specifically trying to make sure that until china will allow equal access to their economy and their country for western firms, we should slow down their ability to acquire and access american high-tech giants and jewels. that's not enough of a strategy, but it's certainly a viable beginning, and that's why i support much of what president trump is doing. what i'll finally say is i'm just going to tick off, and i'm throwing out a lot as a fire hose methodology here for presentation. but i want to just tick off a few of the technologies beyond the ai big data and cyber worlds where i think my co panelists are stronger than i am. where we have to really keep our eye on trends. because these are going to be the areas where much of the other parts of key military competition occur over the next one to three decades. and first of all, anything having to do with nuclear proliferation remains important. not so much for china but for other countries.

so advanced metals, advanced machinery, precision machinery, advanced timing devices, the sorts of things needed to make centrifuges to make bombs themselves, we have to keep a very close eye on these things and not lose sight of that as we try to hasten technology transfer in other domains for good and positive reasons. and then within the areas that i do think are important regarding china in particular, we have to keep our eye on a few things such as submarine quieting technology, which has been a traditional american strength that we've got to try to keep. stealth technology for aircraft, same thing. many aerodynamic and aerospace technologies in advanced engines and hypersonics where china in some ways is getting ahead of us but we have other advantages we can reinforce and preserve. directed energy, including lasers, and then finally, nano materials, microscopic materials, which are very important for everything from batteries to the strength of various composite structural materials we build systems out

of. these are among the areas where i want to interject a note of caution in a conversation that very appropriately is thinking about how do we share more, how do we -- especially in terms of the first panel, how do we promote economic growth through technology transfer between government and private sector. i also want to remind some of the areas where we have to be particularly cautious in my judgment. thanks. >> okay. thank you, mike. so i think we have the headline on the event now. michael han lo michael han long endorses donald trump on some topics. not necessarily every topic. so paul, i want to come to you. so mike mentioned the china connection. and one issue there involves the so-called force tech transfer. where companies claim they have to share their core intellectual property. so i know you work with many companies. so could you provide a perspective from the commercial world on how companies view the intellectual property issue? >> thank you. this is a huge question. and i think the other panelists

have sort of outlined a lot of the themes i would like to touch on briefly. i think i agree with michael on the -- in general on the thrust of some of the actions that have been taken recently. i think it's really important to understand, for example, the 301 action which we're in the middle of now is really not about trade as much as it is about tech transfer. so i recommend reading the ustr port -- it's 215 pages, but the section on tech transfer is very important. tech transfer is mentioned 227 times in that report. and so really, i think we're in the midst of a sort of reassessment of the -- in the u.s. of how we do both export controls and how we handle things like technology transfer. because china has -- for a variety of reasons, has become sort of the poster boy of industrial policy, overreach, in the view of many, including the u.s. government. and things like -- terms like

mercantilism, and it's become a key issue that is really driving in part this trade action. the so-called trade action. and i think there's -- so there's two important parts to that. one is this issue of how do you deal with a country that has a very elaborate set of measures that are designed to compel or -- and in some cases legally involve tech transfer, but there are a whole host of other sort of gray areas, things that are not -- unwritten rules or unwritten documents, and cajoling of companies to do tech transfer and also things like cyber espionage which has been a long standing issue in terms of gaining access to technology. so in part, the process that we embarked in now is an effort to really try to roll back some of china's policies. and it's very -- it's a very complex he had physician that the 301 report calls essentially

china's technology trade -- technology transfer regime. and i think it's going to be a very difficult thing to do, because some of these issues are structural, industrial policies like made in china 2025. things like the national integrated circuit fund, which is a huge fund, which was once described by a u.s. government official as an effort to appropriate the global supply chain for semiconductors. so critical structural issues. what happens is u.s. companies essentially are in the middle of this issue, because u.s. companies are trying to do business in china, trying to negotiate a very difficult regulatory environment to do business there. so a lot of our clients, for example, are involved -- are multinational companies involved in china, and want to include china in their global operations. but have to be very careful how they operate in china. i think the companies we work with by and large view the situation that they can control the amount of tech transfer that happens in their operations in

china, and can protect the sort of secret sauce, if you will, of a company. but that's -- but that, as you'll read -- as you read in the ustr report, you realize how complicated that issue is, because, you know, each company is going to have different problem to deal with, different pressure to do tech transfer. so it becomes very difficult for companies to figure out how to navigate in that market. so in part, there is sort of a mixed feelings about this whole action that the 301 investigation has started. because some companies have been very successful in navigating that and protecting their core intellectual property, and others, of course, have not. so the business community is very split on this, i think. the other piece of this process that will be really important is the investment restrictions. and so there's a major effort in the u.s., of course, to revamp the cifius legislation, the committee for foreign investment in the u.s. and that's a process happening now. and another piece of this whole action against china will be

some sort of proposal of an investment reciprocity regime, which has much bigger implications than the tariff piece. this would restrict chinese investment, for example cloud services, which has become a huge issue in terms of this idea of reciprocity. alibaba and ten cent can build data centers in the u.s. but oracle and microsoft and google and amazon all have to have a joint venture partner and that usually involves some level of tech transfer as part of that deal, for example. so we're embarking now on a really difficult period, i think, in u.s./china relations which encapsulates all of these issues. and at eurasia group, the third top risk was global tech cold war. and i think that the u.s./china cold tech war is a big piece of that. because, again, as we look forward to things like fifth generation mobile, china is going to be a big player in that, and countries in developing markets are going to be looking for technology leaders, for example, for 5g,

and they'll -- the u.s. is looking at how to, for example, build a whole 5g network, what we call china minimized or china-free. so we're -- with tech -- the advanced technology like 5g and particularly ai, there's a lot of sense that this is going to -- we're moving into a world where there's going to be more competition in these areas than collaboration. and i'm also a technology optimist, but a little worried that rhetoric, for example, in the media has tended to focus on the competition and not so much on the collaboration. and for example, in ai, there is a tremendous amount of collaboration right now between china and the u.s. and that's also could be jeopardized by some of these actions coming up. and in any case, the bottom line is the regulatory system is behind on this. wto hasn't worked, and that's one of the reasons we're embarked on this -- the u.s. government action as focused on

301. but we're in for a rough period here, and hopefully at the end of the tunnel, there will be some better sense of how the system can deal with u.s. system and chinese system and global system can deal with these really complex issues related to tech transfer. so i'll stop there. >> okay, great. thank you, paul. so chris, i know you work on social media usage by extremist groups. until recently, people did not think of social media as a sensitive technology, but you show how terrorists have used it to recruit members. there is a similar issue in terms of off the shelf drones. they can be used by hobbyists or terrorists. so do we need to broaden our definition of sensitive technology and if so, how? >> thank you, daryl, for the question, and for including me on the panel. i think a lot of tech transfer debates tend to assume it's between states. and one thing we have discovered over the last decade is really that a lot of the big geopolitical conflicts and

events we have seen have been driven in part by tech transfer to nonstate actors so extremist and terrorist groups and their ability to use technologies for ways their authors and originators never considered. i want to just situate a bit. social media and off-the-shelf drones are uses of commercial technologies. and what we're -- i'm going to mention briefly in a bit, get back to the point about dual use technologies, which is the core issue. but i want to situate how tech transfer even happens to nonstate actors and terrorist groups in the first place. because they don't have the resources to have a big research and development budget. they don't have the resources to acquire cutting-edge technology. so they're really left with three option for getting decent technology. one is the open source movement. they can go on get hub just like the rest of us and download tenser flow and build out their

own sophisticated machine learning models. they can -- the other -- another option is just leeth code. the one thing i'm worried about is what happens when some of the elite cyber weapons the united states has built get into the hands of some pretty bad actors. the third is commercial applications and the cost curve decreasing, more and more become acceptable or accessible to nonstate actors. and the challenge for a nonstate actor is that even though some of them do have technical -- advanced technical capabilities, it's still very hard for them to incorporate new technologies in the same way a state would because they don't have the same level of technical expertise or technical resources. so i was just presenting at the u.n. a few weeks ago on a tech and terrorism conference. and the question was, are terrorist groups going to go down and download tenser flow

and build up their own models for ai to, you know, effectively target u.s. soldiers, for example, in syria. i would be very skeptical of that, because the way ai works. you need a couple -- the algorithm with data. and if you don't have access to massive data and massive compute, as well, it's going to be hard to build your own model. what they can do is take the post train models. post train algorithms that google and others are started releasing for image recognition and incorporate it into an off the shelf drone. and i suspect that's where we're going to start to see this over the next couple years. and i think it's something that we're going to have to pay more and more attention to. the advantage that commercial products have is that they tend to abstract away the complexity of the underlying technology. so if you think about the big app -- social media, we all know about twitter and facebook and their use of social media a few years ago. currently, most of it is

happening on apps like telegram that are end to end encrypted, and what the real breakthrough there is, we have had end to end encryption for a while now. what's new is that you request now have access to it through the smartphone app store. i think people forget what the app store is really doing is abstracting away a lot of the complexity of the underlying technology so it's really just two taps on your phone and suddenly you have access to secure encrypted device that previously really only the pentagon or some other places would have had a decade ago. and so when we talk about the use by nonstate actors of these new technologies, the sensitive technology question, i'm not a lawyer so i don't want to get into the fine details of that, but i do think we need to start thinking very hard about the use of new communications technologies, new robotics, new drones that are commercially available, very cheap. and easy to use.

and think about the dual use nature of them in advance of their product release. i think if you talk to developers at facebook or google today, they would probably admit that they made a mistake five or ten years ago when they set up their platforms and didn't really architect them in a way that would make it difficult for abuse. and i would say that going forward for a lot of the commercial technologies like off the shelf drones, there's a lot we can do to make sure that they are not abused in the same way that isis and other groups abuse twitter and social media. so i have more to say but i'll leave it for there. i think we've already gone over our half hour. >> okay, thank you, chris. i never thought about the app store as a means of technology transfer. but you're right. that is an important point. so i have a question for all of you. and then after this question, we'll open the floor to questions from the audience. so some of you have suggested the need for additional limits on technology transfer. on the first panel, richard

mentioned that 70% of nasa research now is taking place outside of the united states. so the question i want to pose is, if we put new limits on technology transfer, is this going to encourage other countries to do exactly the same thing and with a lot of the r & d taking place outside of the united states, won't this end up harming our ability to innovate? and whoever wants to jump in, feel free to do so. >> i'll start and maybe create the down the row dynamic, because i'll be brief to say i think that's a valid concern which is why i want to target the areas i'm most concerned about into the kinds of groups i mentioned earlier. plus maybe a couple others. but not generalized more than we have to. recognizing if we were to try to do so, we wouldn't be successful in the first instance. we would slow down economic development and growth. and ultimately, you know, it's just not realistic. if you try to limit everything, you're going to limit nothing, because china is too interwoven with the world economy.

so we're going to try to slow china down. it's going to have to be in a number of specific areas. and even there i don't want to slow them down as a matter of permanent policy. i want to force them to play by the rules a little bit and maybe buy us a little more time so their political system matures more before they truly reach our level of super powerdom. >> yeah, i think it's really the question. again, ai provides a good example which we're just sort of grappling with. so, for example, microsoft and google both have hundreds of engineers in china developing ai algorithms. so how are -- is that a u.s. company, chinese company? how do we look at these types of arrangements? and then, of course, chinese companies like alibaba have research institutes in the u.s., hiring u.s. engineers and software developers in the u.s. so that's a -- ai is inherently dual use, too, right?

so there's -- i think we're just coming to grips with that. i think the report that michael mentioned focused particular on ai, and there's a sense now that ai and other things like automation robotics, biotechnology are becoming part in the u.s. of the so-called national security innovation base. and in part, that's what things like the new cfius legislation is designed to better protect. but i think it's going to get complicated, because of these issues of sort of the interactions between communities. and i think, again, ai -- i've done a lot of work on that, and looked at sort of the collaboration between china and the valley, for example. and it's -- it goes very deep. most of china's ai engineers and software developers at their leading companies came through microsoft research in beijing, have close ties to the valley and are very plugged in. there's a lot of chinese investment, as the report points out in startups in the valley

that are driving innovation. so we have to be very careful in developing new ways to protect real national security concerns and assets that we don't stifle innovation inadvertently, particularly in an area like ai, which is still very new in some manner. so the idea of the u.s. government e for example, jumping in and determining through cfius what investments a chinese company can or can't do or vc fund can do in a company in the valley gives a lot of people heartburn, to say the least. so i think these are really valid issues to be grappling with. but i think the danger in extending out some of the -- and revamping the u.s. legal and other measures to control technology is that we end up stifling innovation in key areas.

>> i'll kind of maybe be the debby downer. so it's a good role. i play it often. so i think there's a couple things to think about. and, you know, most of my concerns are about artificial intelligence and related and enabling technologies. one, we have to think clearly not just about ai as a data compute and algorithms, but also kind of the backbone in which ai runs. if we are thinking about gpus, thinking about various types of glitches at the colonel level. if you think about spector and meltdown as indicative of ways of siphoning off technology and other types of secrets and encryptions or whatever you want, these are going to affect worldwide the way in which we can keep secrets and keep things we want to keep secret secret. so just thinking about the stacks and everything from, you know, the chip to the

instruction base to the software to the firmware to, you know, the hardware to everything, right? so i think that's something that we need to take into consideration quite heavily. and then when we think about artificial intelligence, i mean, the sensitive technology thing, you know, i'm looking at this from an application base. and maybe not an investment base. so thinking about applications that are, you know, maybe just not really good ideas. so, for instance, we have the ability right now to generate fake audio, and we can -- so going back to my voice encoding example, i can take any person's voice in this room and get about, you know, 20 hours of data of you talking, and i can create an artificial intelligent agent that just i can make it say anything, and has -- no one can tell the difference if it's your voice or the computer's. so those ai agents can say anything. i can make it say anything. and no one could tell if it's really you or if it's the

computer. not only do we have artificial audio generation, we have artificial video generation. using gans, generaltive adversarial nets. when we start seeing the coupling of things like fake audio and video, anywhere in the world, saying things that could be escalatory or inflammatory and no one can tell the difference if that's the person saying it, that's an application in my view as a weaponization of information. if we want to talk about information operations as an area of conflict as something we have engaged in for decades -- more than that, really. like millennia. but at least we call it information operations for decades. we have to be very careful about those types of technologies that enable those types of military campaigns that are based solely on information. and information in this new --

in this new era, right, is really the heart of it, right? when you can weaponize information, when you can use information to get ahead of your peer. when you can think about ways of using information communication technologies to do this. we have to be very clear about what we're regulating, how we're regulating, and when we think it's crossed a line into weaponization that needs regulation. so, again, i would just kind of walk back a little bit from questions of the structural things, where you can go to school and, you know, where you can invest to just think about the application base, and where you would use that application for any good reason. you know, it might be a fancy little new toggle on my android phone, but is it really necessary, and what are the risks associated with that proliferating to nonstate actors to state actors, to just, you know -- just your angry neighbor down the road, right? so thanks.

>> okay. why don't we open the floor to questions from the audience. so if you have a question, raise your hand. we have microphones and just give us your name and organization. right up here near the front. >> my name is jessica lehman. i recently was working on cfius with the department of defense. my question is protecting u.s. government investments and especially in startups and emerging technology and ai, where there is foreign acquisition, particularly from china, for startups that receive government funding initially. and so how do we address issues of basically u.s. taxpayers funding foreign countries or companies taking over these -- or acquiring these technologies. >> okay. who would like to answer that? and don't everyone speak at once, because that's really rude. >> okay.

i'll take a stab at it. well, i think -- i think on any of these issues like that, the devil is going to be in the details. for example, you know, what sort of government oversight or cfius oversight would be adequate in looking at something as complex as, you know, ten vc companies all have minority investments in a startup that may not have developed a technology that's viable yet, but looks promising. so i think part of the challenge -- and one of the challenges of the dox report and sort of its recommendations is converting that into sort of useful legislation, and then actually enabling a process like cfius to actually make intelligent decisions on this without sort of bouncing the security and the commercial concerns. and my concern is that as structured right now, cfius is heavily on the national security side, obviously, and so may not

have the right personnel or resources to really evaluate some of these more complex challenges that involve sort of earlier stage investment in companies that are doing some cutting-edge technology. so i think there has to be a lot of thought given into how that process works. i think what we'll know -- when we have an example of that, i think the first time that we hear of cfius reviewing, you know, a -- early stage investment in an ai company that involves a chinese minority investor, we'll have a better sense of that. of course, what the reality is that -- and it's mentioned in the report, just raising this issue has already probably served as a deterrent, and potentially discouraged partners or people looking to put together consortium to invest in a particular company from having a chinese partner. so that's already i think probably happened and could happen going forward. but, yeah, these are very difficult questions, and i think

that part of the problem is resourcing properly organizations like cfius to help deal with it. >> i would also say that the devil is in the details, as well, right? so it really depends on how far back you want to go. so if we're talking about, like, you know, early-stage technologies and early-stage companies or companies formed right out of university, right? so if you look at a lot of engineering labs, they're going to come up with some sort of new great new widget, and they're going to patent it. and then they're going to form a company, and probably much of the money that they got to do the research on the widget came from the u.s. government, right? i mean, we look back at google. when you look at the founders of google, right, they took money from the u.s. government through various types of grants. so if we're thinking about sbirs, getting money from darpa, onr, afrl, army research lab. there is so much money from these labs going into university

labs that prop up the kernels of these ideas. and they get to a patentable technology and patent it, and form a really small startup with the lab manager and the guy who invented it. and then they go out and seek vc support to do their startups and they have to think about -- so if you're going from the generation of the idea, which is funded by the united states government, all the way down to vcs investing in a series a, you know, then you have to say, okay, maybe they get over to series a and b and c and all of a sudden by the time you get down to c, you know, you've -- you've got, you know, external investors that you didn't even plan on having in your portfolio to prop up your technology that you don't want to have cost fallacies about, and you start investing -- you start being more open to other types of investors. and that entire kind of patchwork of how an idea gets funded and generated all the way

to when it gets ipoed and thinking about now it's ipoed and now i need a market and a space that needs a market like china or asia, and then i have to have tech transfer. this is such a bee's nest, you know, of -- a beehive, hornet's nest. >> one of those. >> one of those two. >> it's a bad situation. >> it's a bad situation, right? so complex. and the incentive structure that nicole is talking about, if you're a phd student in a lab, you need grants. and the grants that you're going to get that are going to fund you for serious types of -- if i need to build a reactor, i'm not going to get that from the national endowment of humanities or something. i'm going get that from darpa, i'm going to get that from lnr. >> okay. go ahead. >> the one thing i would add to that, building on one of heather's points, it's fundamentally different i think when the tech investment is for a product versus for the talent.

and i think if you look at ai in particular, which you can kind of map out as a function of, again, kind of algorithms, data, compute resources and talent, algorithms are a wash because most are open source. so china and u.s., there's not -- neither of us is going to have an advantage. china probably has an advantage in data. for a whole host of reasons that i won't get into. they also are at parity or maybe pulling ahead in terms of their compute resources. the only advantage that the u.s. really has in this game right now is talent. and so to the extent they're funding our companies to acquire or absorb talent is something that we need to think really hard about. the one example that immediately comes to mind is going from google to bidu. and -- >> stanford to google. >> yeah, i'm sorry, stanford to google to bidu. and i would imagine that he got a fair amount of u.s. government funding through various means

while he was at stanford. so, you know, to complicate it further, i would say we need to not just focus on the technology but the talent itself. >> and on the talent side, with our current interest in cracking down on immigration, it could drive the talent further abroad, which will make this problem much worse. other questions. right here is another question. the gentleman on the aisle. and again, if you can give us your name and organization. >> i'm from china daily. as you may know, icloud service in china has already been transferred to cloud company in a province in far south china. and this company will be responsible for the operations starting from february this year. and i'm just quite curious about your thoughts and opinions on the risk behind it, and what's the consideration of such a deal. thank you. >> yeah, i think that's a

complicated set of business decisions on the part of apple. but that this involved. a couple -- there are two pieces to this. one is the -- the jv requirement in china. apple is essentially operating a cloud service there with icloud. and so they were forced to enhance their local partnership arrangement, and in this case, they close the squajo company. i think also apple is also anticipating some of the provisions under the new cyber security law in china, which are still not finalized, but may require certain companies that are involved in critical information infrastructure provision to localized data. and that would include some foreign companies. although it's still not clear that that's -- that apple would fall under the definition. so i think from commercial point of view, apple made a decision

to also -- because it made sense in terms of customer service and other issues. so i think it was a complicated decision to do that. i think the media has portrayed this as, you know -- as a security issue. apple will be keeping control of the encryption keys for users there, and has said that it will, you know, be very judicious in its legal request. but i think there's a general sense that this is a problem. but i think we haven't really seen an example yet of the chinese government requesting data from apple. but i think the broader issue of law enforcement access to data is part of the whole picture here. so the cloud act was just recently passed in the u.s., which will -- is an attempt to provide a mechanism for law enforcement to gain access to data in the cloud globally. it's going to be very tricky, though, for countries to be

approved by congress as part of the cloud act, and have a bilateral executive relationship so that the law enforcement data can be smoothly passed. so i think -- but your point is well taken. i think there was an earlier panel talked about issues like data localization. in apple's specific case, there were a number of considerations that led to that decision to move to quajo. >> i'm curious how other companies are handling technology transfer. like, are there good examples out there? are there lessons we can learn? good or bad examples. you guys can get back to me in a few minutes if you would like on that. but if you want to pass, we can do that too. >> i want to get back to the topic of not -- of a minute ago on our standing in the world

competitively. and just to add a broader perspective, in addition to the points already made, and this is not in any way to encourage complacency. but chris made a very important point that what we have perhaps over china and other countries now is talent, by which i interpret the entrepreneurial spirit and the people in silicon valley and boston and elsewhere, who are designing new concepts, new applications, new software. and i agree. but it's not just those people, of course. it's the fact that, first of all, they live in the richest country on earth, which also is the center of a western community of more than 1 billion wealthy consumers. and by the way, other people who speak english on this planet include another billion indians who are increasingly wealthy. and much of africa, much of the rest of the world speaks english as a second language. the chinese are a long ways away from being able to compete in these kinds of terms. also, if you were -- if you had an idea to make $10 billion, you would probably prefer to make it in the united states rather than in china, because you have more

confidence in being able to hold on to your own money which gets to nicole's earlier point. maybe we can let these people hold on to their money too much in the united states or maybe we're wealth too much. that's an advantage of a strong legal environment. the competitive advantages of the american economy are profound and we try to erode them through a huge budget feoff sit and a strong washington. not to encourage complacency but to build and put them in a broader context. >> to get away from the consistent discussion with u.s. and china doing this. i think we should also think about, you know, what i think about artificial intelligence is the global spread of talent. if you are looking at major sources of investment and

talent, there's going to be a giant sucking noise going to france. and they are going to be giving incentives for countries to go. you are looking at london as well, the mayor of london said he wants to make london the center of artificial intelligence. you are looking at silicone valley, of course and that doesn't necessarily have to be the place. canada, right? canada is putting so much investment in artificial intelligence. montreal and water lou it is where rim initiated with blackberry. you have toronto being a major hub. if you want to get out of the western world or e.u. side of things. nobody talks about israel.

if you want to talk about cloud computing and robotics, look at israel. they have massive advances in autonomy and a.i. and security applications in particular. i think, again, we need to expand our view outwards from this very narrow western conception as well as the very narrow western conception that is going to be a power play between russia and china. all you did is set the frame and you are going to get blind sided. you're going to be like iran has good computer scientists? yes, they do. >> thank you. that's a good point that we need to broaden the discussion. we are about out of time but we will give nicole the last question. >> nicole turner lee, brookings. i want to bring this

conversation and get your feedback on the facebook/cambridge analytica. universities have been under strict scrutiny when it comes to the development of products and services. developed under iob and the cambridge/analytica scandal was done with the intention of research. it got past alexander and re-purposed along the way and put up this conversation about guardrails. when it comes to the commercial, sector and engagement research. they may not have access to -- they do have access to concealed algorithms so i'm curious,

should we think about stricting scrutiny for the commercial sector engaging in its own terms and feels like permissionless forgiveness. we're sorry, we'll come back and revisit this. again, places, universities and other actors outside of this realm of being innovative in the things they do and puts a security risk. >> that's a great closing question. any thoughts from the panel. >> we need another panel for this one. this is such a great question. the way i think about it is that we do need to push the tech sector to think harder about the negative externalities of what it is doing. i think the risk with facebook,s i view that -- let me back up to say. when we are in d.c. we think of policymaking of the higher level

things that happen in rooms like this. we view governance in terms of political institutions. when we go into digital governance, policy is baked into the level of code and baked into the architecture of how the technologies work. facebook made a decision 10 years ago to effectively growth-hack its platform on how it allows the api to be used and they allowed your friends to have consent over whether or not your data was going to be shared with the third party. they did that because they knew they were going to grow faster and if they do not the, somebody would do it and out-perform them. what we need to be able to do in conversations like this is to begin to communicate to the tech companies, as they are building the problems themselves not ten years down the road but early on

in the product-development cycle to think about what can go wrong here. and you know, there's obviously, we can come up with legislation to target things like what cambridge analytica did. my fear with that is that technology moves so fast, we're going to end up a couple of years ago always. so i think the bigger issue is that folks like yourself, we need to go out to silicone valley and have conversations with them early in the product development cycle. you would have flagged the api choice was a bad idea. it doesn't seem like it was flagged internally. >> if we don't legislate, the europeans will. with gdpr coming up, a lot of people thought if it happened after may 25th that facebook and facebook might be facing action under gdpr which will set the

standard for data privacy and may or may not become a global standard but it will drive regulatory change in europe and influence regulatory change elsewhere. >> to follow up on that nicole. as a fellow at the university of cambridge, the university was not involved in this whatsoever. so just because they co opted the name cambridge, let's make sure you know. my peeps at cambridge are like no, we didn't do that. cambridge analytica, they knew what they were doing. they knowingly broke the law. they were shady enough to know they were breaking the law. and facebook on the other hand, was so negligent in think about whatever i don't know what you are doing, fine, have some data.

i think there was a negligence, gross negligence and gross misconduct and other kinds of things from the platform and knowingly breaking the law. so that is something we have to keep in mind. i would like to flag the work of a friend of mine at washington, saying we do need an fda forolofor algorithms so that when social structures and when it rices to a dual-use or expert controls, we need a federal institution. ryan's work is amazing on this. and then finally, to again, put my hat on as debby downer. they are making everybody kind of a little crazy on how they are going to comply with this cambridge analytica aside. even in the case we might be able to prosecutor them under gdpr but we're not going to be

able to have an arbitration institutions for the average person, right? so gdpr says things that you have a right to your data and look at these things and you have a right to bring -- arbitrate if things become wrong but didn't set up institutions for right of recourse. so you say, like, oh it is great on paper. yeah, you broke the law, i'm going to go to the judge. what is a judge? we don't have the quota on that. unless institutions are created alongside gdpr and not having contradictory things said. gdpr is a good start but i would not hang my hat on this as the regulatory system. mike, the last word on the panel. >> a tiny thought and broader perspective with less knowledge than my co-panel list.

the private giants are being more scrutinized. for 20 years they were the super heros of the modern economy and couldn't walk down the street without celebrating steve jobs and they were creating a new economy and they are amazing people. the edward and snow dens and the government about wikileaks that was taking the scrutiny about what dig data was doing to our lives. the extent of the debate in the early part of the 20th century. and the dot com world, facebook and twitter helping mobile likewising. that was a dynamic for a long time. i doubt the big runs are going to have a by without questions being asked. and the last thing, if you want a general overview of a.i. and

daryl read the new book "the future of work" and it is about the future of work and a nice summary of the other issues we are talking about today and an early plug for your book my friend. >> mike, lunch is on me after that. we are out of time. i want to thank, heather, mike, paul and chris. very enlightening conversation. thank you very much. >> today, fema administrator brock long testifies on the agency 2019 budget. live coverage at 11:00 a.m. eastern on c-span 3. online at c-span.org and on the free c-span radio app. now, a panel of economists kpan the swiss fiscal policy rule, limits the country spending growth to average revenue increases over a