it is very hard to find any organization public or private, by the way, that gets that kind of rope. but you have to give people that kind of rope if you want to go big. prof. miller: one of the under other interesting things about darpa's success or failure comes after decades. i think that point to another challenge about assessing r&d in general, how do you know when it is working? dr. prabhakar: yes. it is not a business where you get to tell on a quarterly basis if you are making your numbers, and the word that gets used a lot is metrics. if you are in the r&d business, everything you can measure mr. going to not tell you much actually -- measure is not going to tell you much actually. number one, you are aiming for huge impact in the future. whatever you are doing r&d, your ultimate impact will be in a future that is too far out for you to manage on a day-to-day basis. and the art of r&d is

for going out how to tell whether or not you are on track to that goal or sometimes a different, even better goal that might open up new prospects, and figure out what can you observe in real-time. when you are managing eight darpa program, what you are looking for is, are the teams coming together and working constructively? are they hitting interim milestones? is the research they are generating really vibrant and exciting and interesting or pretty incremental? that is what you scale by. you stop the things that are not working and very active management and you accelerate others. that is how you work day-to-day. work with the users to persuade them to pick stuff up. famously, they often start off hostile. such a crazy idea. but eventually, they have to go from hostile to at least just skeptical so they will tell you what really has to happen.

eventually like in the mrna vaccine case, a really important moment was when we proved enough that nih said we need to partner with moderna to start moving into the real world. that is what those steps look like. again, we are talking about darpa, but president biden started this for health, recognizing that kind of high-impact orientation was something we need to bring to the challenge of health outcomes for americans. and taking those methodologies and adapting them to a very different set of issues, different science base, different technology issues, very different ways to achieve impact. but i think there are many elements of that model with energy have been successfully adopted, which is great to see. prof. miller: i guess that

points to the question of how easily we can scale our r&d structure. you called for spending more. are we convinced it would be useful? dr. prabhakar: i think we are in a deficit right now so we have to do some fixing up. the thing that is really boring to everyone but has to happen, which is making sure we have facilities and infrastructure that work. i go out to the ni t where i was director 30 years ago and i was so proud because it represented the scientific progress we had. it was under instruction for a long time because they cannot finish the construction because we cannot seem to get construction dollars appropriated to keeping the facilities going. it is also many other places. we have to do so many foundational things. it is time for us to go back.

the possibilities that science and technology really bring us. i will use ai as an example because today it is the most powerful technology that is shaping opportunities. this administration has done a lot of work to get ai on the right track. the president and vice president seized the generative ai moment that started about a year and a half ago, a couple years ago, just saying we have to manage its risks so that we can -- excuse me -- so we can harness its benefits. we have done a lot to get on a better track in terms of managing risks of safety and security and bias and discrimination. and i think there is more to be done there. but i have really been eager to turn our focus now to the second part, which is how are we going to use this to go big? when i talk to my colleagues in the tech industry, they for sure

are looking at every possible business opportunity on productivity and enhancing the way we do all kinds of work. as long as they do that responsibly, that will be great for the country. i want to see that happen, but there are things we need to use ai for that are not just for things the market is going to drive. they are for public missions doing the country's work. i think about how we design and approve drugs. i am thinking about how we close educational gaps for our kids, which we have been trying to do for a long time and have not made the kind of progress we need. i am talking about how we deliver our weather forecast. think about what is going on right this minute with the climate, the exacerbated weather events. think about our transportation infrastructure. think about how the government delivers its services to its citizens. we can see these tantalizing prospects about how ai can

transform each of those areas, but those are all going to take serious investment in r&d. and i think to me that points to the reason to get federal r&d on a healthy track is not just to compete. it is not just to r&d make sure we keep the base healthy. it is so we can do these things. prof. miller: let's open it to questions for a couple minutes. i would look to hear your thoughts on how you think about that very large area and what the priorities are when it comes to r&d. dr. prabhakar: the term biotechnology for decades now has meant pharmaceuticals, the new way to build drugs that are really effective. there we have an industry that is well underway and there are

critical supply-chain issues that need to be looked at for the health of that industry but the core technology of using biology to synthesize chemicals and materials is now something as most capabilities have expanded. there are interesting prospects building on pharmaceuticals that are really tantalizing. one of those at darpa, we had a synthetic biology program. the program manager came in and said there is a future where the air force instead of pouring a runway could grow a runway. a biological process that harvests materials in this place you just landed in and actually can build something of that structural capacity. that is actually something we have proven you can do now. there is an effort to build that out. this sort of gives you a sense of how much more biology can do and where we are today. this was the focus of an

executive order the president signed a couple years ago now in order to make sure that america leads in this area. what i don't want to do here is a playbook we have done many times which is we do beautiful research and someone else commercializes it and we do not have the manufacturing capacity here. one of the areas we have focused on is building the bio process caleb capacity so that -- scale up capacity so that we not only know how to do things in a lie but we can start showing what scale up looks like to vent out the process so it can then be commercialized much more rapidly. prof. miller: an example of the need to combine r&d efforts and supply chain analysis at the same time. dr. prabhakar: the whole picture. that is exactly right. prof. miller: let me see if we have any questions from the audience. please, sir. >> hi, good afternoon. i have a quick question. i joined late but i heard you talk about health and combining ai.

so i wanted to apply it to cancer. it seems like that has been the long-running issue, and there are so many different types of cancer. we still don't have something beyond the traditional chemotherapy and immunotherapy to circumvent the various cancers. can we apply it to that? i think that would be a celebratory day. dr. prabhakar: i with you. president biden started it when he was vice president, and he and the first lady reignited it a couple years ago. you know this is personal for them having lost a son to cancer, but cancer is personal to every person in america and around the world who has been touched by this horrible problem. and the goal they set when they reignited the cancer moonshot, the quantitative goal was double the rate of progress we have been making and to cut the cancer death rate in half over 25 years.

that would be 4 million lives saved from cancer deaths in that time period. that is a huge deal. i have been thrilled to see daniel carnaval, who is part of my team at ostp, the leader of the cancer moonshot. thrilled to see her approach, not just as an area for more research, but an area where we are putting in place regulations that will prevent new cases of cancer. for example, from pfas and other contaminants. doing the work of screening and early detection so we can nip cancer from the bud. americans missed 10 million cancer screenings during 2020 because of the pandemic. we have regained that and gotten screening out to many people, including people who did not have health insurance, did not have primary care. we are making progress on that. and therapies, making sure the therapies that exist are affordable and reached by more people, but also that the new therapies can come to the fore.

very much to your point about using a height, there are so many possibilities. i will mention one program that is using ai to look at approved therapies and discerning which of those might be good candidates for rare diseases, including rare cancers, which are very difficult, very challenging to get therapies for. but mapping a safe and effective treatment for other diseases. looks like a very high potential way to accelerate treatment for rare diseases. >> hi. claude with aei. in decades past, the washington policy community, the science community has obsessed with the role of the science advisor and the relationship with the president and the rest of the apparatus. how has that role changed? if you look back at the bushes

or clinton or even obama, how has the role of the science advisor changed, if it has, over the past decades? dr. prabhakar: i find that the role of the science advisor is really shaped by the president, number one. i had the great privilege of coming into this role, working for a president who has always seen science and technology as -- this is a president who loves to talk about how america is a country that can be defined in a single word, and that word is technologies. technology. he has always been a supporter for science and technology for its myriad of applications. how it changes health outcomes to how it changes economic opportunity. part of being a great nation is understanding how the universe was shaped and pursuing the

boundaries of fundamental knowledge. he has always appreciated all of this, so that made it great fun to be in this role. a lot of my time in this administer should has been shaped by the fact that chatgpt showed up about a month after i showed up in this job so a lot of my direct interactions with the president have focused on artificial intelligence including showing him chatgpt and image generators early on like a year and a half ago. that was a session that kicked off a lot of the ai work. and because of his approach, both he and the vice president really sought ai as bringing growth promise and peril. they have approached it from a perspective of being aggressive about going after risks but also making sure that we are laying the groundwork so we can seize the opportunity. that has been one of the areas of tremendous focus in my

interactions with him. >> thank you. thank you for both being here. fabulous discussion. chris i hope you choose to write a follow-up to chip war, perhaps chip war 2.0. dr. prabhakar: what do you think? >> speaking of semiconductors, i am curious. because of the chips and science act, a lot of the billions of dollars have been allocated to intel. intel, as you know, has had some challenges. i am just kind of curious, if you were working with intel or could offer advice to intel, what kind of advice and guidance might you provide to have intel challenge other types of semiconductor companies?

dr. prabhakar: i don't think i am in a good position to give advice to intel. if you look at what the chips act is getting done, i mentioned the important shift from 0% of advanced node manufacturing in the u.s. to getting to 28% over a decade. what that represents is not one single company, but now we are on track to have the five leading-edge manufacturers from around the world have operations, manufacturing production operations in the united states. that to me is really significant, that cohort as a group. no other nation has more than two. and putting us on a path where we have five of those gives us the position that we are seeking to get to. prof. miller: in the back? >> a research public policy

professor at george washington university and the last remaining member of the carter administration's industrial policy sleeper cell. dr. prabhakar: wow. >> because in 19 80 the carter administration was in the process of coming up with an industrial strategy that would have been implemented except for the election. so i got in the business of doing industrial strategy at the regional level, probably 200 projects, and the question has to do with the collaboration between the public sector and the private sector and how that gets done, because right now the chips act, the way the government is used to working now is through rules. the chips act is all about the rules. to be eligible for this, you need to meet the following criteria. there is nothing about business and government collaboration, and the high point of the collaboration, maybe chris knows this, was under the esteemed

commerce secretary in the 1920's, herbert hoover, who had industrial strategy and set up mechanisms for business and government collaboration that happened to be white male anglo-saxon protestants so they were comfortable working with each other. but the two mechanisms where boards, the remnants of which is in the commerce department now, the international trade administration. and the other was the commerce department setting up trade associations, beginning with the u.s. chamber of commerce. the u.s. chamber of commerce was created in 1912 by the federal government to be a partner for industrial policy with the federal government. that changed during the new deal. so this is all a question too, how do we set up mechanisms that are transparent and accountable in which government and business come to some mind meld because it is the private sector that produces the stuff, not the public sector? dr. prabhakar: i love that history.

that was great. my view is our economy and the look we are trying to do, it is not single threaded. it is complex, and you have to break down the goals as in figure out what the nature of the partnership needs to be. so in the case of chips, we were very clear that the country's goal is to make sure we have the advanced semiconductor manufacturing we need sufficiently located in the u.s. and we have our allied relationships in a way that meets our economic goals and our military goals. that is what the country is trying to get done. of course, every partnership, every arrangement that secretary raimondo's commerce department is putting together with each company, the company is we together because they have business goals and they find enough shared interest to make something happen. that is the nature of those

partnerships. it is a very different kind of partnership when you are doing the kind of r&d i was describing in my darpa story. i think we need to be open to many different kinds of relationships and arrangements, but all goal driven and recognizing what the partnership represents is the government acting on behalf of what the country needs and the companies acting on behalf of what they need because they are entities that have their own goal. i would not look for a cookie cutter. i would like to have many, many different approaches. >> so word on the street is there is an election coming up. dr. prabhakar: how about that? >> it is kind of hard to call so your time at ostp may or may not be coming to an end in the next few months. kind of a three-part question. any of which you may be willing

that you would say we are so proud to achieve this? number two, anything at the opposite end where you say i wish i could have done this but because of unforeseen obstacles? and number three, what are you looking forward to doing next? dr. prabhakar: i have no idea on the third one. i think about one and two all the time. that is the day job, figuring out if you made it. i am proud of what this administration has gotten done to put ai on the right track. tons more work to be done. but ignoring this powerful technological course that is we shaping society was not a good plan and the fact that we have been active there is something i am very, very proud of. more globally, i think something that has been a great passion of mine and a reason i was so excited to have this opportunity is that federal r&d, in my view,

is still too much on auto drive from and this frontier, which is now 80 years ago -- endless frontier, which is 80 years ago. it is too long without really rethinking the roles and purposes and the way to accomplish those roles and purposes for the federal r&d enterprise. and i would tell you i think we made very important progress in, for example, starting to focus on health outcomes and starting with the new technology innovation partnerships director at nsf on the important basic research we do and making sure we are aiming the whole enterprise at achieving. look, the whole contract between the american people and federal r&d is when r&d will deliver a better tomorrow for the american people, and ethic we have gotten a little better at that. at that is a very hard thing to

do so i am proud of the progress we have made their. the thing that i regret that i think we have but that this country simply has to do, we have got to go big. we are not taking many big shots at really huge opportunities. part of the reason i am excited about ai is it is disrupting every area, but if you figure out how to use it and wheeled it, it can be an enormous accelerator. i think that is true for our most important societal goals that we talked about, opening up access to opportunity for every person, economic growth, national security goals, health, dealing with the climate crisis. i think the opportunities are immense. i think this r&d enterprise we have in this country, public and private, is fully capable, but

we have got to aim it at resources to generally achieve the big things i know are possible. prof. miller: could i follow their? making sure we are going big and achieving goals, what is the limiting factor? you mentioned research and aiming. those are important to him about people would agree going big is better than doing small, so why are we not? dr. prabhakar: let's talk about what going big would look like. a great example is when the senate ai leadership group that majority leader schumer put together, they came out with their ai roadmap a while ago. one of the things they mentioned was a $2 billion resource for nondefense ai r&d. i think a lot of people looked at it and said, $32 billion, what would you use that for? to put that in context, as a government we spent close to $200 billion a year on r&d.

32 billion dollars is a 13% kicker to that. if you give them $32 billion on top of what we are currently doing in r&d, that will take federal r&d from about 7% of gdp to 7.7% of gdp, so it starts moving you in the direction of sufficient federal r&d investment to meet our ambitions. what does it look like on the ground? what it looks like is today we have thousands of known diseases for which there is no treatment, and we generate new pharmaceuticals and address new conditions at the rate of 20 or 30 a year, so thousands, 20 to 30. if you can speed that up tenfold or a hundredfold, think about what that means. that means my kids might be able to grow up without worrying about dementia at the end of their lives. that is just unbelievable but it could be possible. it really could be possible.

we are doing pieces of the basic research to pursue that. pharma companies are each taking their own trove of private data and trying to make progress on that. but until we pull all of that together and supercharge it with investment, my kids are not going to achieve that future. i want them to have that future. we got work to do. one or two prof. miller: --prof. miller: one or two final questions, please. our two final questions, and then we will wrap up after that. >> thank you so much. i am with the m.i.t. washington office and my question would be the influence on china. our national r&d. looking at it from an m.i.t. perspective, the international students harnessing the innovation possibilities that their on premise brings but also managing the risks that might

appear with ip. china high-risk on the watchlist, how do you see the future or what r&d has to do what has to be done? dr. prabhakar: this is an issue that has been very challenging in the research enterprise and especially universities. it is not that long ago that everyone was being encouraged to build these linkages with chinese research organizations. and it started to recognize that china's behavior has changed over the last 10 or 15 years and they have crossed one boundary after another, whether it is militarily in economic interchanges or going after intellectual property. in research in particular, i think how you deal with that has been a vexing issue for you. it has been a vexing issue for all of us as well. and we recently put out security

guidance to try to wrangle three facts. one of china's aggressive and inappropriate behavior. the second is the fact that research, basic research can only thrive with open global communications. you cannot lose that in the process of being less naive about what china is doing. the third factor is at the individual level, which is a core value in this country is we do not discriminate against people because of their ethnic background, and that cannot be part of how we deal with this situation. i think it is -- we are getting guidance in place i hope will make it easier for universities to, you know, be just a little smarter about how to deal with these issues while making sure we treat people right and keep that openness that is so vital for research. >> thank you so much. >> i am from the american

university. thank you so much for this discussion today. you already talked about something that does not exist which is a big step, but i am concerned about the things more upstream of the steps. so do you feel like there is an attention being paid to that? thank you. dr. prabhakar: i'm sorry. do you mean for example materials? >> yes. at the end, the production going in for the chemicals to be used. my concern is at some point we build all of these great things but do we have the supplies to run them? dr. prabhakar: first of all, critical minerals and critical materials have to be paid attention to. i don't want to be five or 10 years down the road saying i wish we had done something back

in the mid-2020's. that is a great example of an area we need to be looking to the future. i mentioned the part of the chips r&d effort going on. building on top of what we were talking about with respect to what ai can do. the pressures on advanced materials for the next generation of semiconductors are extraordinary. they are trying to do technically very challenging things to make chips like. they have to worry about supply chains. and a lot of those materials and processes use pfas and other chemicals that are very problematic as we learn more about the health implications, and the regulatory pinch is definitely affecting semiconductor manufacturing today. that seems like an almost unsolvable problem, but to me it is a perfect example of where really innovative research can make a difference. so this chips r&d effort recently launched an initiative

that is about using ai first to expand the universe's possible materials and combinations of materials and then coupling that with autonomous experimentation to much more rapidly drive practical solutions that can be transitioned into production much more quickly. i love that because when i see an impossible problem, i am itching to find science and technology ways to make that turn into a trackable problem. if we can do that, there are just in manufacturing remaining and continuing to build in the u.s. which is part of the r&d effort. prof. miller: it has been a fascinating conversation. dr. prabhakar: great to be with you. [applause]