Biodata as an instrument of National Power

Transcript

hi I'm Craig Smith a former New York Times correspondent and host of the podcast I on [Music] AI I'm also a special government employee at the National Security Commission on artificial intelligence and in this role I'm serving as the host for nsse ai's podcast series on the commission's work this is the fifth of sixth episodes looking at the commission's third quarter recommendation to Congress this week we talked to Chris Darby an nscai commissioner and CEO of inel the CIA back tech Investment Company about the commission's recommendations that AI for National Security be extended to biotechnology and that biod dat be considered an instrument of National Power I hope you find the conversation as fascinating as I did [Music] Chris generally We Begin by having you introduce yourself give us some of your background what you're doing at inel and how you came to the commission sure well first of all Craig thank you very much let me start by saying what an honor it is to be serving on this commission alongside such an esteemed group led by Eric Schmidt and Bob work and of course sap by an incredibly dedicated talented group of Staff supporting the commission it's been a real pleasure thus far I came to the commission appointed by CER Warner my day job I've been with a company called incel for the last 14 years as the CEO were the investment arm of the United States intelligence Community doing early stage investing prior to that I grew up in technology primarily in the Telecommunications and Computing space you must have followed Gilman is that right was he the CEO before you at inq wasn't he CEO at inel Gilman was early at inel I did not follow Gilman directly into the job I see your line of effort begins this quarter's recommendations with a call to use AI for large scale biotechnology research and Innovation I've spoken to people before about Ai and biotechnology but I thought maybe you could talk a little bit about how AI will drive new Innovations in biotech and why did the commission choose to take a specific in-depth look at Ai and biotechnology sure well AI is really all about its application to adjacent Technologies and with bio I think we're probably seeing the the beginning of the next big Tech Revolution it really begins with the fact that biologies become programmable whether it's plants animals humans we're getting to the point where biology can be treated as a set of ones and zeros that can be edited the same way computer program can be edited and so AI is really the key to enhancing our understanding of these core biological processes and enabling transformational breakthroughs and Craig really the power of this biological intersection set with things like AI can impact Health Food environment all of these things going forward can really change the way we live on a day-to-day basis really by some estimates it's an incredibly powerful tool some numbers would suggest that 60% of physical inputs to the economy will be produced with sin bio going forward and so the US has really got a maintain leadership in biotech and ensure that biotech and AI strategies are really mutually reinforcing yeah can you talk a little bit about the 60% of physical inputs in the economy coming from synthetic biology just give an example for listeners so they can visualize that sure well we're dealing with situations today where we're trying to optimize crops for production with less water for example and so you can engineer a crop where it is actually consuming less water than it once did and of course there are great environmental reasons for wanting to do things like this we are seeing production of actually materials based on biology there are companies out there I believe gko bioworks in Boston doing really some Leading Edge things about taking for example yeast and using that as a baseline to generate other products s that's fascinating I've had Alan asaru and some other people on the podcast talking about new materials through biotech you also call for reducing Reliance on Chinese entities for large-scale genomic research databases and you recommend widening the definition of biological threats in the National biod defense strategy to include biodata advantages as an instrument of National Power can you talk about that threat it sounds as though you're referring to something specific well I think if we start with the code of life being programmable and data being the new oil it's all about how much of this data can you amass we've got to go into that with our eyes wide open biotech is dual use and there's a dark side to the potential of the intersection set of AI applications to bio and the commission sites three main threats one strategic and really two are operational the Strategic threat is that bio Tech is an instrument of National Power it can be a projection of power and for example withholding access to either key Technologies or drugs could impact another nation's State uh Strategic Benefit could be gained through the economic growth derived from the foundational bioinnovations and really you could see a situation where one nation was advantaged for example in the area of Health to the detriment of everyone else in the world we just need to understand that it is at the Strategic level an instrument and a projection of the power of of a Nation at the more tactical operational level really I think there are two things that the commission concerns itself with one would be engineered pathogens we've seen I think Craig the devastation that can be caused economically and physically from an organic pandemic can you imagine if something was engineered for lethality it just is a concerning situation when the keep me up at night much more even than the nuclear threat I think that it's something that we as mankind will need to deal with and then there are somewhat lesser areas areas like human enhancement should we engineer better humans stronger humans taller humans that sort of area is an area that is a concern for the commission because there are ethical and value conversations that have to go on so I think what the commission is recommending is really that we have to recognize that China's Ai and biotech strategies are mutually reinforcing and and of course they have very good reasons that I referred to earlier for investigating these areas and moving forward in these areas but there is a dark side so we're recommending updating the national biod defense strategy to include biorelated threats that can be enabled with AI we just think it's The Prudent thing for us to do from a national security perspective yeah and I understand the Confluence of biological threats and AI there are man-made viruses you can synthesize a virus in the lab for example just on the first point of the economic advantage or the medical advantage of having biologically engineered drugs that could help one country or give one country an advantage over another it sounds like you're talking a little bit about the hunt for the vaccine with covid-19 did that inform this recommend a at all I don't think vaccine production specifically informed the recommendation as much as it was a macro level recognition that the next tech Revolution is going to be bio and that the intersection set of AI and bio will transform economies and societies going forward and I think the commission rightly said that we have to look at the application of AI in the area of bio as being an instrument of nation state power of National Security concern and really of the concern of mankind going forward and and so when we get into questions like who's value systems and so on are being applied to these things we need to take it very seriously and using the term biodata as opposed to biosynthesis or some other term why is data the focus as opposed to processes or outcomes well I think in Ai and generally speaking in the data Sciences if you have more and diverse data sets you can train your algorithms and you can gain insights at a higher Fidelity than you can with less data SPS data is just more challenging and so to the extent that you have more data you have more opportunity and if you have diverse data you have a wider aperture of things that you can be addressing and so we believe that data is really the foundation upon which you apply your AI and to the extent that you can amass large amounts of that data gives you Advantage you know the other thing I would say is that we traditionally talked about biodata within the context of genomics and I think as sequencing got cheap the genomic collection got aggressive by some people and we're equally seeing now that for example things like brain waves and patterns of brain waves could be a data set that is collected so anything that is being derived from a human being or an animal or a plant can be collected and then tools can be applied to edit or understand those targets and when you say include biodata in the National biod defense strategy you're talking about preventing the collection of biodata from American citizens or Allies or protecting the biodata of American citizens or Allies well I think what we're talking about is first of all the ethical use of biodata and I think to the extent that there are laws around privacy that we believe in here in the United States and that our allies believe in I think we've got to understand what those laws say and what our general societal tolerances for our genetic makeup to be made available for people who have these intents and one of the things that I think we're most worried about is just the fact that a lot of China's work in these areas and that the bgi in particular the Beijing genomics Institute what they're doing with this massive data set is just opaque to us we don't know we do know that they're collecting both domestically aggressively and around the world now and the information they're collecting around the world is going back to China as I said before these data sets are dual use we want a healthier population we know that the costs associated with Healthcare going forward are going to be high here in the United States and in China and so to the extent that we can understand the genome and we can create a healthier world I think that's a good thing but as I said earlier there's a dark side to all of this too and I think transparency is important in these areas and I think standards are and value systems are important in these areas and this has got to be a conversation that goes on I think the commission recognizes that and is just trying to highlight that it starts with [Music] data on bgi can you talk just a little bit more about who they are and what kind of research they're engaged in the biodata collection that they're engaged in sure the Beijing genomics Institute is really the de facto Chinese national champion in data sequencing and research they're the big dog in terms of what China is doing on the biodata and research perspective we don't have a lot of clarity on what they're doing but here's what we do know we know that the Chinese government is interested in collecting massive amounts of data and making a database of DNA info for example on I think theyve got 700 million Chinese men in the data set we know that they're also collecting internationally and of course the covid-19 pandemic has given them an opportunity to put facilities around the world and collect in the name of trying to be of assistance on the pandemic but BG also runs China's national Gene bank and so when we look at that collection we think well okay beyond the pandemic what are they doing with it we also know that they're subject to China's cyber security laws that require them to share data with the intelligence and National Security apparatus and I think that of course that whole family of laws and the intersection set between China's commercial interests and the interests of their intelligence and National Security apparatus gives the commission pause we know that bgi has received for example substantial loans from China's government and not surprisingly there's some evidence that suggests that they work reasonably closely with the pla so bgi may be serving wittingly or unwittingly as a global collection mechanism for China's government and enhancing that National genetic data set that they could use in a variety of different ways yeah there's a whole conversation to be had about biodata and its potential uses the commission recommends announcing a set of specific use cases for quantum computers to help spur private sector investment and Innovation and transforming Quantum Technologies can you give some examples of possible use cases specifically for National Security sure and let me start by saying I'm not a Quantum expert to the extent that I am on the commission I've been listening to the people that are much smarter than I am on on these subjects the private sector works best when given a challenge the private sector loves to innovate around the challenge and if you look at the application of AI and Quantum the US military's got big challenges that could be addressed and so things like supply chain operations could potentially be optimized using Quantum and pre-positioning of different resources or where they need to be prepositioned could be anticipated using Quantum more quickly and in a probably more accurate way than classical computing we also know that quantum computers could be used to simulate atomic reactions inside a nuclear reactor on the power grid or maybe even a nuclear weapon but the fact is that we need to just throw out some of these different areas and let people iterate on them let them see where Quantum can be applied where it will work where it won't work Quantum is still in its infancy and I think the best way to get the private sector involved is to give them a couple of challenges and say optimize my supply chain show me how this is going to work and I've got a lot of faith in the United States creative Innovation engine I think they will come up with things that we can't even imagine as Commissioners yeah it just seems a little early because Quantum Computing has really only been shown to work theoretically I mean the Google Quantum Supremacy exercise you know has created a lot of controversy about what exactly that means to apply Quantum Computing to specific problems I mean you have to first make a quantum computer that can compute and you recommend making Quantum Computing accessible to researchers via the national AI research resource we've talked about the resource before but maybe you can explain again to listeners that haven't heard previous episodes what the national AI research resource is and then talk about why you think it's not premature to talk about making Quantum Computing access accessible or setting challenges for the private sector well Craig let me flip the order here let me start by saying why it's important to get involved now I think that whether we're talking about gate based Quantum or adiabatic Quantum I think that the technology is evolving and it's really about the number of error corrected cubits that any particular platform are delivering at one point in time and arguably you need to get to approximately 100 eror corrected cubits before things get really interesting in Quantum but that said there's C C sensitivity to that estimation so what could we do with a lesser number than a 100 of error corrected cubits and I think that's really the question that we as a commission are trying to encourage people to explore and so the national AI research resource the nare that the commission recommended in q1 would really provide researchers with access to compute resources government data sets Educational Tools support to try and begin to interrogate the existing quantum computers such as they are they're not at the level of airor corrected cubits that we believe you need to be really relevant going forward on some of these problems but we might be wrong there may be things that we can do at hybrid Computing environments where you combine classical with Quantum Resources at a lesser than 100 Cubit level to make inroads and so really our basic message is that the United States needs to lead basic Quantum res research but it risks falling behind in applying this Quantum Computing to the National Security use cases if we are so focused on a really Exquisite use case and if we delay in giving these creative people access and it's a recurring message I think with me we have to enable the creativity of our private sector any way we can and I think the N is really a platform or a resource to do that [Music] the commission says that the United States should take steps now to cement its long-term status as a global leader in the design and manufacturing of quantum processing units what kind of funding are you talking about what size of funding and how does that compare to funding currently available yeah I think Craig the the thing that the commission's really hitting on here is that we don't want to see a situation going forward that mirrors the microelectronics conversation that we're having to have nationally right now where we largely lost our advantage in production to other nations around the world where once we had it in the basic microelectronics space we don't want to see that same thing occur in Quantum and you're right the private sector and the government has invested heavily in Quantum on the research side and the development side I think we do lead probably in the design area and Manufacturing and we want to maintain that lead So to avoid making the mistakes for quantum computers that we did for semiconductors we've got to ensure that all our future advances in design and manufacturing of quantum processing units occur here in the United States and in terms of the dollars we really are recommending that Congress create a package of Provisions that incentivize the domestic design and Manufacturing here in the United States with the goals of keep a really vibrant ecosystem and in terms of the levers that Congress has well I think the commission really talks to tax credits for companies that are seeking to access quantum computers for business Ops and potentially things like loan guarantees could be utilized and I would suggest even thoughtful Equity financing for specific startups and companies that are doing unique things to advant adust in the quantum space I don't think it's about a specific size at this point in time as much as it is we have a number of tools in the tool bag and we should bring them to bear on this challenge does d-wave which is sort of the dominant manufacturer at this point of quantum Computing systems of early quantum computers do they get government funding or tax breaks or something like that to encourage their research well full disclosure inel is an investor at d-wave and so to the extent that inel invests on behalf of the United States national security Community yes dwa has gotten money from the US government but they're not the only Quantum Computing company out there they're companies like retti IQ the US has got a vibrant Quantum startup community and I think we've got to nurture that your line of effort recommends incentivizing private sector development of a domestic Foundry and this is for classical semiconductors a Foundry costs as much as 20 billion I've had a conversation about this with Gilman l another commissioner what kind of funding does the commission recommend for a Foundry I think that we got to start by saying why is this important if you look at the United States right now the Innovation the small company Innovation are really based on fabulous semi companies so you can think of them as design firms and right now these design firms in the microelectronics space have to go to Merchant fabrication facilities offshore the two dominant ones being tsmc and Samsung and so really if you start with our goal our goal is really to bring that Merchant fabrication capabilities back here to the United States to really turbocharge our startup community and our broader research and development community in the microelectronic space the commission is recommending tax incentives rather than publicly funding a foundary so to us we want to encourage others to build the foundary or to bring a found to the United States a 40% tax credit on semi manufacturing equipment and facilities in the US through 20124 would probably do the trick and incentivize the private sector construction of a merchant fabrication facility here in the US specifically tax incentives that were included in the chips for America Bill introduced by Senators Warner and Corin over the summer would do the trick and really I think that those sorts of things are very important but they're not included in the NDA right now that's under consideration on Capitol Hill and I think that's probably something that should be Revisited the chips for America bill I think was well reasoned specifically around turbocharging the Innovation engine in the United States and assuring that we have a lead going forward in the microelectronics space we as a commission also recommended public R&D funding for micro Electronics through DARPA I think it makes great sense their electronic Resurgence initiative is a really really good one the National Science found A's got some efforts going on and of course doe has got significant thought leadership in these areas so we looked at a number of different places where incentives could be put in place and funding could go to really make sure that this notion of a merchant Fab capacity spawning Innovation was resonant here in the United States and with those credits I mean Samsung and tsmc are part of the backbone of our alliances with South Korea and Taiwan would those credits apply as well to foreign manufacturers if they're building foundaries in the United States I know tsmc is building a plant in Arizona a Foundry I believe and would that be taken into consideration that if you're reshoring so much of the microelectronic supply chain you're not cutting off or hurting the Allies that have relied on that expertise to build their economies yeah well I don't think it's a one or the other situation I think if you look at the conversations going on with psmc they're predicated on the chips for America Act and the tax incentives that would be put in place that make it viably economic for them to as you put it resure to the United States we are not for a second suggesting that they would need to close their plants in Taiwan or in South Korea if it were Samsung we're suggesting that here in the United States we have a substantial Market in of ourselves to consume the capacity that they would deliver here in the US and just makes good sense for them to if the economics can be balanced correctly put their facilities here in the US yeah and I wasn't suggesting that they would close there I'm just saying that a domestic us Foundry would Supply a portion of the US market that otherwise has to buy from Taiwan or South Korea but then again the demand for semiconductors is expanding exponentially right now it seems your line of effort also makes several recommendations for tracing Supply chains and reshoring critical elements in the microelectronics manufacturing chain would that include pushing for the domestic manufacturer of extreme ultraviolet lithography equipment for which the US now depends on the Netherlands I believe I think that we have to understand that the supply chains for semiconduct Condors the tool sets and so on are highly globalized the globalized nature of the semiconductor industry really prohibits you from trying to reshore every single element of it as you rightly point out we are dependent on the Netherlands for some of the euv technology the lithography equipment that is so crucial and I think that's fine I think that it would be imprudent of us to try and reshore every element of the supply chain especially where it's a point piece of the supply chain that we can consume from an ally I think what we've got to do is we've got to look at sort of the critical elements and more importantly where they're going to advance over a period of time one of the things that the commission has recognized is that for example packaging is going to be very important going forward in addition to the next generation of etching or deposition tools that may evolve but the overall packaging environment is one that will differentiate semiconductors going forward and so what we're suggesting is that tax credits and subsidies incentivize the US firms that are really going to push the envelope on some of these new areas while still purchasing from our allies where the Allies have clear leads and differentiated products so the lithography equipment I mean that's another huge investment to develop that technology domestically so you're not talking specifically about the lith graphy equipment that may remain with our allies well the beauty of tech Craig is that it moves fast and so I think what I'd be asking is what's the next generation of tool going to look like how's the tools Market going to get disrupted going forward we want to incentivize us startups and someone to try and do that and I think that that is where the best bang for our buck could come that's interesting it sounds like you're talking about something specific with regard to etching are you well I think etching deposition Tech are areas that are right for disruption when you look at the way microelectronics could evolve we're seeing things like chiplets which are little tiny processors that you put together in unique ways we're seeing conversations about stackable Technologies in other words layering things vertically as opposed to Simply horizontally and as you get into these new form factors it's going to require new tools and frankly whole new ways of envisioning fabrication going forward the US has to do what the US does best we innovate well and we've got to recognize that today our allies can provide unique capabilities and they innovate well in certain areas but we've got to push the envelope I think it was Wayne Gretzky's dad said skate to where the puck is going not to where the puck is today and I think that's what we've got to try and do and I think that's what the commission is trying to [Music] incentivized the commission recommends empowering a single entity within the White House to strengthen executive leadership and Tech policy where does that leadership reside now and is the recommendation meant simply to raise the profile of tech in policy debates well this one's near dear to my heart Craig as somebody that's growing up in Tech in the private sector we have a national security policy we have a National Economic Policy we have no National technology policy and if the world goes the way I suspect it will and biotech for example is the next Revolution we've got to craft a technology vision for this country and embark on an all of nation approach to realize that vision and and so my question is who in the United States owns that you've got the the National Economic Community you've got the National Security Community you've got ostp but each deal with different parts of the US technology policy question and none of that really is coordinated from a tech perspective and I would ask where does it all come together where do we maximize our investments in a more unified way recognizing the adjacency that I referred to earlier AI in and of itself is really only valuable as you talk to the adjacent technologies that it is going to disrupt and so Tech is Central to strategic competition that we will have going forward and you can't Silo it and so what the commission's really recommending is a tech competitiveness Council to be chaired by the Vice President and to be led by a newly created assistant to the president to really make sure that these different very valuable entities doing Tech work in the United States come together in a line on a single or at least on an agreed to Tech vision for the country that sounds very much like the American Technology Council that already exists but failed to hold the interest of the Tech Community if it's in the office of the executive given the polarized nature of our politics today how do you insulate that from politics maybe it should reside someplace other than the executive is what I'm suggesting well I would suggest that when things get done in Washington they get done at the executive level because time is of the essence this is not a problem that we can afford to stare at for decades as a nation and we've seen that a BP level Council we saw it in space with NASA it can be very very effective and I think it will be up to the vice president to make sure that the council's rep represented of a variety of different interests in political viewpoints especially as we get into things like biotech where clearly privacy and human rights and so on have a very significant role to play in the conversation but we basically have to treat Tech as a national security concern and as a national opportunity if we can solve for health if we can solve for food if we can solve for environment utilizing something like biotech the United un States should lead the world in that our population will be better off for it and so will the global population look these are serious times with serious challenges as we've seen with this pandemic we can't afford for politics or anything else to get in the way of us leading and solving these challenges so the commission's recommendations I think are a good start this biod data thing really fascinates me because I understand the biowarfare threat but I really hadn't thought of biodata as being the base for that threat it's a threat and it's an opportunity and and one of the things that Craig I'm trying to balance here is as we've seen National Security is much more than kinetic event the fact that wuwei is a proxy for aircraft characters now being deployed around the world and some time is really the payload that they're delivering neither of those things is kinetic and so bio's got to be viewed very similarly it is something that can change lives and it's also something that could be used against us if someone so desired Yeah Yeah Chris I do hope we have an opportunity to talk again another time because this stuff is endlessly [Music] interesting that's it for this week's podcast I want to thank Chris for his time if you want to learn more about the National Security Commission on AI visit their website at www.nc.gov and remember the singularity may not be near but AI is about to change your world so pay attention [Music]