Transcripts For CSPAN2 Hearing 20240704 : vimarsana.com

Half. The subcommittee will come to order and i ask for unanimous consent to share be authorized to declare a recess at any time. Without objection so ordered. I want to briefly review the three commandments of the subcommittee. One is that we shall start on time which we just did so thats good. We will enforce the five minute rule that if time allows we will entertain a second round of questions, and it often does. But i bring this up i want to stress the third which is shall not use acronyms or jargon that is particularly important in this discussion because the discussion about ai can quickly generate were degenerate into jargon laden discussions with three true experts on the topic but dont assume your average member of congress or dont understand i dont assume i understand the nuances when you get into ai. We want to have a discussion with your average american to understand today. We are asking you to demystify a lot of the concepts surrounding aia and the thing about the topic that may sound counterintuitive but ive been going back to the history of the early cold war in particular i am obsessed with the korean war in which the moment in which the cold war first turned very hot and at great cost to americans and even greater cost. I was reading a book about it and came across the words that said a part of the west military thought plays in the belief that machines must be used to save its mens lives. Korea would progressively become a in illustration of the effects of the limited war where one side possesses the firepower into the other than manpower. Theres a lot of ways to interpret this in the context of this hearing one is that it can potentially increase the destructive power of modern war variant together as it has the potential to decrease it or at least decrease exposure that the sailors, airmen put themselves in a combat situation or the third of that is unique in contrast to the early cold war warriors the machines themselves might somehow take power and go beyond our ability to control them. Today we want to dig into these different hypotheses. I have dug into this topic and i want to commend the Ranking Member for the way and which hes worked with me to really use the subcommittee to explore the concept and we have a very fascinating discussion with elon musk last week. There were some sources of disagreement. He believes china is on Team Humanity and the only thing ive become convinced the ccp if they win the competition or the sort of component of the competition we will likely use that technology for evil as a way of perfecting the repressive totalitarian surveillance state as well as exploring that model around the world whereas we in the free world at least have a chance of using it for good. To make sense of these things we are lucky to have incredible witnesses. Mr. Alex wang the ceo of skill ai and it might be the most successful mit dropout at this point but theres probably a unique set of sub people that would qualify. Mr. Kitchen as a senior fellow at the American Enterprise institute and is someone many of us on capitol hill look for for advice when talking about the intersection of technology and warfare. And doctor Haniyeh Mahmoudian of data robot as well. Ive been looking forward to the hearing fordo a long time and lk forward to an honest and open discussion just remember no acronyms and jargon and with that i will yield to the Ranking Member. Thank you mr. Chairman and thank you for convening the panel and your interest in a bipartisan way in addressing ai aching sure that our military is leading with ai. I appreciate how you approached this. Im not going to be long because people want to hear from the witnesses. I would just say my understanding is that china is spending almost ten times as much as then u. S. As a of ther military budget on ai and we needto to think about the modern technologies that are going to be needed to have the most Effective National Security Strategy so i am particularly curious from the witnesses about how they think that america can contain and have the lead in the Ai Technology Going Forward. What are the investments we need to make and what are the standards we need to have to ensure that our ai is used most effectively. Im looking forward to the panel. Thank you. Mr. Wang you are recognized for five minutes. Chairman gallagher, Ranking Members of the subcommittee, i amit the founder and ceo of scae ai. Its an honor to be here to testify. One thatof will be dominated byi end of the United States must do to win. In 2016 i founded a scale with a mission to accelerate the development of ai from the earliest days of working with of the leading Autonomous Vehicle programs at General Motors and toyota, Technology Companies such as meta, microsoft and open ai and partnerships with the u. S. Government including the u. S. Department of defense, the u. S. Army and u. S. Air force. And at the forefront of Ai Development for more than seven years. The country must integrate new technology and war fighting to win. If we dont win on ai we risk global democracy to strategic adversaries like china. The National Security mission is deeply personal for me. I grew up in the shadow of a nationalist up. My parents were physicists and worked on the technology that defined the last era of warfare the atomic bomb. The Chinese Communist party deeply understands thehe potentl for ai to disrupt were variant is investing to capitalize on the opportunity. I saw this firsthand four years ago when i went to an investor trip to china both enlightening and unsettling. China was making rapid progress developing the technologies like facial recognition and using these for domestic surveillance and repression. The same year their president said, quote, we must ensure the country marches in the front ranks when it comes to Theoretical Research innt this area of ai and occupy the high ground in critical Ai Technologies. A china is investing the full power t of its Industrial Base r ai. This year roughly three times the u. S. Government on ai. Its also heavily investing in the enabled drones down to Radar Systems, Autonomous Vehicles and has launcheded over 79 models since 2020. Ai is chinas apollo project. To lead the world in the development of ai we must lead highquality data powering ai. Scales committed to doing our part to support the u. S. Government to ensure america remains at a strategic advantage. Today we do so in three ways. One, scale data. We prepare data for the u. S. Government. Autonomous Mission Systems we partner to develop a data engine to support the vehicle program. Three, we develop the ai power decisionmaking platform that rapidly helps u. S. Government to sense of realworld information. The dod has also takenor a numbr of steps in the right direction most notably with the launch of an Artificial Intelligence office. While the progress is promising more must be done to ensure. The over matches our life pillar to plan to maintain the United States security and technological edge in this era. First, investment in ai. Its critical to increase American Investment to maintain the leadership despite the investment in the 24 budget the u. S. Is still spending three times less than china. A second the data supremacy. The systems are only as good as the data they are trained on. The dod creates more than 22 terabytes most of which is wasted. Ai warfare is leading the world in developing the data scaling supporting the legislative mandate to establish thet repository that enables the dod to harness the data. Third, testing and evaluation. Its one of the most important ways to ensure the models are accurate, reliable and uphold the ethical goals. The administration embraced this by highlighting an Evaluation Platform for frontier and death. Fourth, pathfinder project as congress should Fund Programs with a mission of developing innovative capabilities. Since the project started more than six years ago no new pathfinder projects have begun. Fifth, upscaling the workforce the u. S. Should invest in rapidly training the workforce for ai. They already worked on this with the dod to tackle the challenge head on. In st. Louis we established a center that created more than 300 ranging from entrylevel to Machine Learning engineers with advanced degrees. The race for the Global Leadership is well underway and i couldnt be more excited to do everythingng in my power to ense the u. S. Wins. Its moments like this congress, the dod and the Tech Industry can rise to the challenge together or stand idle. Ive included my further remarks in a statement to be submitted for the record and thank you for the opportunity to be here today. Mr. Kitchen you are recognized. Good morning members of the committee, thank you for the privilege of testifying. Id like to use my Opening Statement to make three points. First, i believe Artificial Intelligence in particular ai and National Security lifeline for the United States. The community discussed the potential of ai for years but now it seems they are finally maturing and can be applied at a scale with few doubting that they will soon reshape almost every aspect of our lives gincluding how we fight and win wars. At the importance is felt as acutely in beijing as it is in washington but until recently i wasnt confident the United States would hold the ai advantage. If you assume this advantage comes down to algorithms, data and hardware just one year ago i wouldd have given the United States the advantage on algorithms the chinesee the advantage on data and called hardware but this deserves anotherju look. Large language models and other ai may be moving it to the americanan advantage. The u. S. Dominates the Computer Science giving birth to these advancements and we remained the choice for Global Talent. On the hardware a strong bipartisan consensus is allowing us to meaningfully constrain access to cuttingedge capabilities like advanced graphics processing units and even more can and should be done for example limiting Chinese Cloud Services would be an excellent next step. Finally on data while the chinese economy and people continue to generate digitized data and of the Chinese Communist Party Continues to have unfettered access to the data the promise of Synthetic Data and the fact many of the new models are indexed on the open internet made blunt the advantage. It is my hope for example that the Chinese Governments political strict content controls and general oppression of its own people will compromise or bias matches the data it collects diluting its utility and ultimately limiting the development of chinese ai. At the very least i think the United States has an opportunity to surge ahead if we are aggressive andsi deliberate. But it offers the u. S. More than this capability. Large language models and other technologies ifte properly realized couldld provide an economic base for the new era of prosperity and security. For years weve known the United States is not investing sufficiently to meet the demands of the nation. The truth of this has been laid fair as the industrial race struggles to keep upri with the demand of the conflict in ukraine for example but according to one recent study at existing aia capabilities could add to the equivalent of 2. 6 trillion to 4. 4 trillion annually to the Global Economy and this estimate would double if we included the impact of the existingto software that is currently used. The bottom line is this. I believe ai is offering an opportunity to get our economic house in order to lay a foundation for the nations longterm prosperity and to build a National Security enterprise that is properly resourced. But finally, while ai offers all of this promise and more it also hass a serious National Security risks. Most acutely a flood of misinformation and the exponential growth of conventional and novel cyber attacks. By now we have all seen the photos, videos ands, other media ai is creating into the capabilities have already been democratized. Virtually everyone can create and distribute synthetic media that will undoubtedly be used to undermine confidence and our democratic institutions. Similarly they will offer hostile cyber actors potent tools for generating and automating traditional and new online attacks. In the world that we are already overwhelmed by online threats, it will soon for gas on these fires. S. Theres much more i could say on these matters but i trust we will cover them over the course of the hearing. Thank you again for the opportunity to testify and i look forward to your questions. Thank you. Doctor, you are recognized for five minutes. Thank you for the opportunity to testify before the subcommittee on the Critical Issues off Machine Learning and human warfare, Artificial Intelligence on the battlefield. I am at global ai. I am a member to the National Advisory committee and coach the working group. Today i testify in my digital capacity. Ai holds immense potential and is increasingly becoming an essential component of modern military strategies and operations with a potential to profoundly impact Operational Efficiency ander decisionmakin. In the realm of cybersecurity it can help to protect the network and systems against increasingly sophisticated Cyber Threats and also a system on cyber operations. Ai can also play a Critical Role in predicting and the prevention of injuries among military personnel. It can efficiently track in injury which can aid the prevention of other bodily injuries which with these consequences is a major reason for medical disability and service. Thus it isha imperative for the United States to expedite the adoption to sustain our advantage. It is crucial to ensure it adheres to the law and ethical guidelines. In recent years the comprehension of ai potential adverse effects have left to numerous instances despite the good intentions ended up arming individuals and groups it was designed to help. This suggests the consideration often relegated with of the building and deploying the ai systems. However it is encouraging that the department has taken initiatives to develop these principles that apply to both combat and noncombat functions. Ai technology will change much about the battlefield of the future but nothing will change the steadfast commitment to responsible and lawful behavior. It requires consideration of people, the process and technology. The literacy from all levels accused them to ensure responsible and appropriate use of ai. To successfully adopt ai at scale in the department of defense requires that the department implement the Governance Framework and adopt Risk Management processes to manage and mitigate the risk associated with ai. One of the challenges and adoption of the government especially in this department of defense is a slow procurement process. As mentioned earlier,o it is an evolving space, therefore it is paramount for us to make sure that we have a faster procurement cycle, but the proper violation of the ai tools by using robust processes. In conclusion, ai holds transformative potential. However, along these benefits, it is vital to establish ethical frameworks and comprehensive government processes that ensures effectiveness, reliability and human oversight. Thank you. Thank you for the thoughtful testimony. I recognize myself for five minutes. I like to begin by asking you to respond a bit to some of what mr. Kitchen laid out in terms of our advantages and disadvantages relative to china in the ai race. How do you see those advantages and disadvantages . I certainly agreeee that its the place of choice so we certainly continue the advantage and the evidence is clear as well as the model that underpins it and all of those were invented in the United States. When it comes to data i agree that we have a potential when it pertains to military implementations. In america we have the largest fleet of military hardware in world. This fleet generates 22 terabytes every day so if we can properly set up an instrument, this data that is being generated into pools of ai ready data sets then we can create a pretty insurmountable to military use we need to Work Together and move towards the country. Today most of this data is wasted in some manner and we need to fix that to create a longstanding and durable advantage in Ranking Member<\/a> for the way and which hes worked with me to really use the subcommittee to explore the concept and we have a very fascinating discussion with elon musk last week. There were some sources of disagreement. He believes china is on Team Humanity<\/a> and the only thing ive become convinced the ccp if they win the competition or the sort of component of the competition we will likely use that technology for evil as a way of perfecting the repressive totalitarian surveillance state as well as exploring that model around the world whereas we in the free world at least have a chance of using it for good. To make sense of these things we are lucky to have incredible witnesses. Mr. Alex wang the ceo of skill ai and it might be the most successful mit dropout at this point but theres probably a unique set of sub people that would qualify. Mr. Kitchen as a senior fellow at the American Enterprise<\/a> institute and is someone many of us on capitol hill look for for advice when talking about the intersection of technology and warfare. And doctor Haniyeh Mahmoudian<\/a> of data robot as well. Ive been looking forward to the hearing fordo a long time and lk forward to an honest and open discussion just remember no acronyms and jargon and with that i will yield to the Ranking Member<\/a>. Thank you mr. Chairman and thank you for convening the panel and your interest in a bipartisan way in addressing ai aching sure that our military is leading with ai. I appreciate how you approached this. Im not going to be long because people want to hear from the witnesses. I would just say my understanding is that china is spending almost ten times as much as then u. S. As a of ther military budget on ai and we needto to think about the modern technologies that are going to be needed to have the most Effective National<\/a> Security Strategy<\/a> so i am particularly curious from the witnesses about how they think that america can contain and have the lead in the Ai Technology<\/a> Going Forward<\/a>. What are the investments we need to make and what are the standards we need to have to ensure that our ai is used most effectively. Im looking forward to the panel. Thank you. Mr. Wang you are recognized for five minutes. Chairman gallagher, Ranking Member<\/a>s of the subcommittee, i amit the founder and ceo of scae ai. Its an honor to be here to testify. One thatof will be dominated byi end of the United States<\/a> must do to win. In 2016 i founded a scale with a mission to accelerate the development of ai from the earliest days of working with of the leading Autonomous Vehicle<\/a> programs at General Motors<\/a> and toyota, Technology Companies<\/a> such as meta, microsoft and open ai and partnerships with the u. S. Government including the u. S. Department of defense, the u. S. Army and u. S. Air force. And at the forefront of Ai Development<\/a> for more than seven years. The country must integrate new technology and war fighting to win. If we dont win on ai we risk global democracy to strategic adversaries like china. The National Security<\/a> mission is deeply personal for me. I grew up in the shadow of a nationalist up. My parents were physicists and worked on the technology that defined the last era of warfare the atomic bomb. The Chinese Communist<\/a> party deeply understands thehe potentl for ai to disrupt were variant is investing to capitalize on the opportunity. I saw this firsthand four years ago when i went to an investor trip to china both enlightening and unsettling. China was making rapid progress developing the technologies like facial recognition and using these for domestic surveillance and repression. The same year their president said, quote, we must ensure the country marches in the front ranks when it comes to Theoretical Research<\/a> innt this area of ai and occupy the high ground in critical Ai Technologies<\/a>. A china is investing the full power t of its Industrial Base<\/a> r ai. This year roughly three times the u. S. Government on ai. Its also heavily investing in the enabled drones down to Radar Systems<\/a>, Autonomous Vehicle<\/a>s and has launcheded over 79 models since 2020. Ai is chinas apollo project. To lead the world in the development of ai we must lead highquality data powering ai. Scales committed to doing our part to support the u. S. Government to ensure america remains at a strategic advantage. Today we do so in three ways. One, scale data. We prepare data for the u. S. Government. Autonomous Mission Systems<\/a> we partner to develop a data engine to support the vehicle program. Three, we develop the ai power decisionmaking platform that rapidly helps u. S. Government to sense of realworld information. The dod has also takenor a numbr of steps in the right direction most notably with the launch of an Artificial Intelligence<\/a> office. While the progress is promising more must be done to ensure. The over matches our life pillar to plan to maintain the United States<\/a> security and technological edge in this era. First, investment in ai. Its critical to increase American Investment<\/a> to maintain the leadership despite the investment in the 24 budget the u. S. Is still spending three times less than china. A second the data supremacy. The systems are only as good as the data they are trained on. The dod creates more than 22 terabytes most of which is wasted. Ai warfare is leading the world in developing the data scaling supporting the legislative mandate to establish thet repository that enables the dod to harness the data. Third, testing and evaluation. Its one of the most important ways to ensure the models are accurate, reliable and uphold the ethical goals. The administration embraced this by highlighting an Evaluation Platform<\/a> for frontier and death. Fourth, pathfinder project as congress should Fund Programs<\/a> with a mission of developing innovative capabilities. Since the project started more than six years ago no new pathfinder projects have begun. Fifth, upscaling the workforce the u. S. Should invest in rapidly training the workforce for ai. They already worked on this with the dod to tackle the challenge head on. In st. Louis we established a center that created more than 300 ranging from entrylevel to Machine Learning<\/a> engineers with advanced degrees. The race for the Global Leadership<\/a> is well underway and i couldnt be more excited to do everythingng in my power to ense the u. S. Wins. Its moments like this congress, the dod and the Tech Industry<\/a> can rise to the challenge together or stand idle. Ive included my further remarks in a statement to be submitted for the record and thank you for the opportunity to be here today. Mr. Kitchen you are recognized. Good morning members of the committee, thank you for the privilege of testifying. Id like to use my Opening Statement<\/a> to make three points. First, i believe Artificial Intelligence<\/a> in particular ai and National Security<\/a> lifeline for the United States<\/a>. The community discussed the potential of ai for years but now it seems they are finally maturing and can be applied at a scale with few doubting that they will soon reshape almost every aspect of our lives gincluding how we fight and win wars. At the importance is felt as acutely in beijing as it is in washington but until recently i wasnt confident the United States<\/a> would hold the ai advantage. If you assume this advantage comes down to algorithms, data and hardware just one year ago i wouldd have given the United States<\/a> the advantage on algorithms the chinesee the advantage on data and called hardware but this deserves anotherju look. Large language models and other ai may be moving it to the americanan advantage. The u. S. Dominates the Computer Science<\/a> giving birth to these advancements and we remained the choice for Global Talent<\/a>. On the hardware a strong bipartisan consensus is allowing us to meaningfully constrain access to cuttingedge capabilities like advanced graphics processing units and even more can and should be done for example limiting Chinese Cloud Services<\/a> would be an excellent next step. Finally on data while the chinese economy and people continue to generate digitized data and of the Chinese Communist<\/a> Party Continues<\/a> to have unfettered access to the data the promise of Synthetic Data<\/a> and the fact many of the new models are indexed on the open internet made blunt the advantage. It is my hope for example that the Chinese Government<\/a>s political strict content controls and general oppression of its own people will compromise or bias matches the data it collects diluting its utility and ultimately limiting the development of chinese ai. At the very least i think the United States<\/a> has an opportunity to surge ahead if we are aggressive andsi deliberate. But it offers the u. S. More than this capability. Large language models and other technologies ifte properly realized couldld provide an economic base for the new era of prosperity and security. For years weve known the United States<\/a> is not investing sufficiently to meet the demands of the nation. The truth of this has been laid fair as the industrial race struggles to keep upri with the demand of the conflict in ukraine for example but according to one recent study at existing aia capabilities could add to the equivalent of 2. 6 trillion to 4. 4 trillion annually to the Global Economy<\/a> and this estimate would double if we included the impact of the existingto software that is currently used. The bottom line is this. I believe ai is offering an opportunity to get our economic house in order to lay a foundation for the nations longterm prosperity and to build a National Security<\/a> enterprise that is properly resourced. But finally, while ai offers all of this promise and more it also hass a serious National Security<\/a> risks. Most acutely a flood of misinformation and the exponential growth of conventional and novel cyber attacks. By now we have all seen the photos, videos ands, other media ai is creating into the capabilities have already been democratized. Virtually everyone can create and distribute synthetic media that will undoubtedly be used to undermine confidence and our democratic institutions. Similarly they will offer hostile cyber actors potent tools for generating and automating traditional and new online attacks. In the world that we are already overwhelmed by online threats, it will soon for gas on these fires. S. Theres much more i could say on these matters but i trust we will cover them over the course of the hearing. Thank you again for the opportunity to testify and i look forward to your questions. Thank you. Doctor, you are recognized for five minutes. Thank you for the opportunity to testify before the subcommittee on the Critical Issues<\/a> off Machine Learning<\/a> and human warfare, Artificial Intelligence<\/a> on the battlefield. I am at global ai. I am a member to the National Advisory<\/a> committee and coach the working group. Today i testify in my digital capacity. Ai holds immense potential and is increasingly becoming an essential component of modern military strategies and operations with a potential to profoundly impact Operational Efficiency<\/a> ander decisionmakin. In the realm of cybersecurity it can help to protect the network and systems against increasingly sophisticated Cyber Threats<\/a> and also a system on cyber operations. Ai can also play a Critical Role<\/a> in predicting and the prevention of injuries among military personnel. It can efficiently track in injury which can aid the prevention of other bodily injuries which with these consequences is a major reason for medical disability and service. Thus it isha imperative for the United States<\/a> to expedite the adoption to sustain our advantage. It is crucial to ensure it adheres to the law and ethical guidelines. In recent years the comprehension of ai potential adverse effects have left to numerous instances despite the good intentions ended up arming individuals and groups it was designed to help. This suggests the consideration often relegated with of the building and deploying the ai systems. However it is encouraging that the department has taken initiatives to develop these principles that apply to both combat and noncombat functions. Ai technology will change much about the battlefield of the future but nothing will change the steadfast commitment to responsible and lawful behavior. It requires consideration of people, the process and technology. The literacy from all levels accused them to ensure responsible and appropriate use of ai. To successfully adopt ai at scale in the department of defense requires that the department implement the Governance Framework<\/a> and adopt Risk Management<\/a> processes to manage and mitigate the risk associated with ai. One of the challenges and adoption of the government especially in this department of defense is a slow procurement process. As mentioned earlier,o it is an evolving space, therefore it is paramount for us to make sure that we have a faster procurement cycle, but the proper violation of the ai tools by using robust processes. In conclusion, ai holds transformative potential. However, along these benefits, it is vital to establish ethical frameworks and comprehensive government processes that ensures effectiveness, reliability and human oversight. Thank you. Thank you for the thoughtful testimony. I recognize myself for five minutes. I like to begin by asking you to respond a bit to some of what mr. Kitchen laid out in terms of our advantages and disadvantages relative to china in the ai race. How do you see those advantages and disadvantages . I certainly agreeee that its the place of choice so we certainly continue the advantage and the evidence is clear as well as the model that underpins it and all of those were invented in the United States<\/a>. When it comes to data i agree that we have a potential when it pertains to military implementations. In america we have the largest fleet of military hardware in world. This fleet generates 22 terabytes every day so if we can properly set up an instrument, this data that is being generated into pools of ai ready data sets then we can create a pretty insurmountable to military use we need to Work Together<\/a> and move towards the country. Today most of this data is wasted in some manner and we need to fix that to create a longstanding and durable advantage in Artificial Intelligence<\/a> data and when it the computer usual power and video which is the leader in the trips for Artificial Intelligence<\/a> as an american company. These technologies are innovative and built in america. So there i think we have an advantage. You delta what with the pentagon. Why is it at present that it is wasted . What iss preventing us from harnessing that data and what is preventing us, why have we not had the new pathfinder project assistance maven . So, the data is something that is significantly more valuable with the amendment of f these Artificial Intelligence<\/a> algorithms, so a very simplistic way is that theyve analyzed troves and patterns of data and can emulate those patterns Going Forward<\/a> so we view that with models like chat that are able to read the troves of data and written over years and years and then it can emulate the human rights begin. So these Artificial Intelligence<\/a> gatherings and algorithms have valuable more than they have in the past and so it isan a new paradigm that the dod needs to adopt. As we all know it is a fragmented organization. There is many different constituencies and organizations that each have their own approach to data and like one of my witnesses mentioned theres an education process and Everyone Needs<\/a> to understand thatly its the ammunition in te war and if we have that recognition as an entire department and as a country i cothink it becomes clear to take the right actions to collect the data and move forward. It sounds as if you are suggesting with the right leadership and organization the dod can actually be a leader in this. And of the guardrails a lot of our constituents are asking us about. With of the ethical framework is that a foundation that could be built upon expanded to end sure we are on the same page within the alliance and gradually bring more and more people into that framework . I definitely agree i think it is critical that the United States<\/a> takes the lead on this topic particularly as it pertains to ensuring Artificial Intelligence<\/a> is used in accordance with our values and principles. The dod has established principles that i believe are great and thosew principles are ones we should continue and i think now comes down to implementation. How we are going to make sure themel principles are followed d that is what i think that the regime is incredibly important and critical to increase the deployment of the systems. As the dod looks to apply to every function within its operation from war fighting to logistics we need to have proper test h evaluation mechanisms to ensure that every Artificial Intelligence<\/a> deployed follows our ethical ai principles. So we need to make sure it follows these principles and folead the world in terms of thinking how it can be used in accordance with democratic principles. I recognizeco mr. Connor for five minutes. Thank you mr. Chairman. Three quick questions, quick answers i hope. Mr. Kitchen mentioned Global Talent<\/a> but we also have immigration issues here hindering. Indeed or should we change some of the barriers that exist to get that a Global Talent<\/a> here to the u. S. To make sure we are not losing that talent to other countries . Maintaining the access and continuing to be the home of the talent would be essential for National Security<\/a>. Its fragmented. Could you give the committee as a followup someme suggestions n procurement changes within the area of ai is that something that can be carved out. I agree that its a major problem. But can we do something specifically . One area that we think about and i have to emphasize the military is not my area of expertise, but one area that i can bring from the business perspective. It is a long process of evaluation with standard procedures in place it would be this type of evaluation as long as they are standardized it can go much faster. You are familiar though on side. Litary can you follow up on that . Inse building the faster procurement methods for the department of defense, notably the chief Digital Ai Office<\/a> set up a trade wind program that is one of the fastest procurement methods for new and Innovative Technologies<\/a> as an organization its also been actively partnering with many Innovative Tech Companies<\/a> and bringing them into the dod. So there are current programs i think we can doubled down on. Both of these instances i mentionedd are working and need to look towards the next era doubling down on some of these best procurement methods and ensuring we continue. In the two minutes that i have left, it was mentioned that we have a data advantage in the u. S. I think inherently in our democracy with privacy rights protections we are at a disadvantage in terms of how chinese operate themselves and it cant just be broken down to information military and otherwise. All that information is valuable. Is there an area because of the privacy that is changed. China views these technologies and they are likely going to squash. We have protection of free speech and we will continue innovating when it comes to the Large Language Models<\/a> and they view that as a risk to be the socialist values. Ii never looked at that aspect f the answer. Lastly with 30 seconds to go, putin has said whoever controls ai has a huge advantage but look at russia right now. Is it fair to say that they are way behind and is it fair to say that they are involved in ukraine and what itss doing to the economy and the sanctions having an impact yes or no . Fourteen seconds. Theres good reason to suspect the russian ai capability while they may w have basic research in terms of applied deployment is minimal. I think the Ranking Member<\/a> for switching his time so i can go to another hearing. Doctor mccormick is recognized. Thank you mr. Chair. And to the witnesses, i wish i had time to talk with you all day because this is fascinating and you are experts. We have time enough for about so i will go with the most pertinent that you brought up in your Opening Statement<\/a>s. You just discussed limiting chinese access to the information, which totally makes sense we see how. They develop when they literally take the confirmation and apply it. We are right back there. That is access to leading technology in america. Is that what youre discussing when you talk about access or are you talking about in the industry itself. Certainly there is undeniable level of risk associated with of the student presence in the United States<\/a> however the research that ive seen by the organizations like the Georgetown Center<\/a> for the security demonstrates that the vast majority of the Foreign Research<\/a> students even Chinese Students<\/a> stay in the United States<\/a> or more broadly in the west for the course of their career. And amplify our capability. But i am most concerned about, however, when i talk about the access to data again, not dismissing the inherent to build in threat there is frankly the acquisition through purchase of americanrc data through the lare data stores but theres things weve been talking aboutav and staring in the face for multiple years now thingsth like chinese owned and operated a social Media Companies<\/a> like tiktok. It is by law made accessible to the Chinese Communist<\/a> party. So while Chinese Students<\/a> and other foreign students may have some type of risk, it pales in my view in comparison to the type of data that we are just kind of giving away. I appreciate that and i can totally understand where thats coming from. My otherot concern though in regards to that and a quick comment, the Chinese Government<\/a> is not stupid and they obviously dont care about their people more than their government so when they allow people to come in with education for jobs, i think it is with a various intent and that is my worry im not saying we need to educate people but im worried about it and about anybody that is pushing people over here knowing they are not coming back for a reason. The one thing that we do have a huge advantage is people investing in the aia and they dont have the capacity to outperform because they dont have a private industry. How do we compare when we provide our synergistic efforts between the government and private industry where china as far as and you mentioned the effect o that we allow this freedom to flow and its not controlled so it does have a potential to outpace as long as we put the right guardrails on itit when we are talking about e competition with china. If you factor in the amount of private Sector Investment<\/a> in the a United States<\/a> that is a an incredible song. Large Technology Companies<\/a> into the Venture Capital<\/a> industry and now the sort of Global Enterprises<\/a> investing billions and billions of dollars into so if you tally all of that up its an incredible investment into Artificial Intelligence<\/a> in the United States<\/a>. That being said i dont think that we should rest easy on that because its going to be incredibly important. O we need to engine were in the next phase that the u. S. Both economically dominant but also has military leadership as well whenip it comes to Artificial Intelligence<\/a>. So we need to consider the Overall Investment<\/a> and thats where theres a large disparity investing more and if we compare that percentage of the overall military investment they are spending between one to 2 of their overall budget into Artificial Intelligence<\/a> where the dod is spending between. 1 and. 2 . It is interesting to watch the private industry is now in the United States<\/a> pairing with the dod to develop a lot of this stuff. I will say since im out of time that we shouldnt sleep on iran and russia who obviously want to be players that he was technologies to disrupt other countries and of course they love misinformation so this is something. With that iof will yield. I thought it was interesting that you said the advantage china may have because of data is diminishing because of things like chat are based on the universe of the internet that has both good and bad data. And then mr. Wang, you said the dod is relying on i the sort of tagged and annotated data. I guess im trying to understand what is the best data that is neededne for aia to be effective in the military applications and does china have an advantage of that kind of data or not, and i would love both of your answers on that. Both the data are important and both open source data that is accessible on the internet is a key data source for chat gtp as well as highquality annotated data sets and the precursor were trained on large quantities of highquality expert generated data to ensure the systems are more trustworthy, truthful et cetera. So, both matter, but when we look towards again military implementations of ai, the key is what is the military data that includes. Right now the models that are used by consumers and the president of the private sector are trained on essentially no military data. Towards the data problems they would not perform particularly lwell. We need to have military aia data sets that are ready for this kind of deployment. When it comes to that kind of data i think probably today you would say its a jump ball. Looking at this issue and the dod is looking at this issue but we have all of the fundamentals to having an insurmountable advantage because the dod generates the terabytes of data far more than on a daily basis. So if we can instrument the data into one central repository, we can come out ahead. Being a Surveillance Data<\/a> from other citizens wasnt going to be helpful for the military data set that is needed for some military problems. It will be a very limited help. At the data is orders of magnitude more valuable. I think the application matters. Anything that could be tactical or kinetic the military generated curated data is going to be the key differential. The point that i was trying to raise when i mentioned the data advantage, its been once aspirational. The hope of the generative ai would be Synthetic Data<\/a> so instead of data thats been produced with normal Economic Activity<\/a> or military Economic Activity<\/a> they are able to then begin generating Synthetic Data<\/a> sets that would be useful for training. I suspect number one we are not there yet and number two that those will be helpful for broad economic applications but not at all the type where they will be supplemental for the applications that alex wasas discussing. Doctor mahmoudian, thank you for your testimony. I know in the framework guide length, im not sure if thats been updated now. Are there things you want to do more in terms of the ethical guidelines and framework for the use of ai . As i mentioned they already have the Ethics Principles<\/a> in place. One comment that i would have about that is how we can make these frameworks to practical. That comes with the education of the personnel to make sure they understand what these principles mean and how they can impact this to the cases that they are working on. So that is the first step. The second step is the implementation of the Ai Governance<\/a>, so when you are talking about policies and processes that the government would have, those measurements that would be part of the process would include the principles that they have. So its all about people into the process and obviously the technology. How we are going to measure the risks these are all part of the technology aspects. An explanationde of why the sysm make certain decisions. Thank you for bringing in the extent we have to think about all these Weapon System<\/a>s that we havent contested and the in theenvironment as the data collectionsrm platforms. What advice would you have for the committee about the access or utilization regime for the data that we are currently wasting . I think this is one of the most important things we can do to set up america for decades and decades of leadership and military use of ai. Right now a lot of this data goes onto hard drives and what ends up happening are the hard drives are eitherr overwritten with information so the old data gets deleted and effectively and lost or they go intoiv sort of closets or places where they never see the light of day. First it is instrumenting the data to sort of flow into one central Data Repository<\/a>. They have a legislative mandate to do so and set up for the dod so i think that is of critical importance and then this is a whole love dod issue. Every service, every group, every Program Needs<\/a> to be thinkinghe about how can all the data other programs are collecting andnd generating witn their purview how can they am surell they flow through into oe central repository and then are prepared and tagged and labeled for the ready usehe down the li. It would seem as though under the normal construct of emissions that someone might reasonably be stovepipe to waive from the broad utilization of the data so it almost seems like something that is an appendage to a mission set, very hard to leave it in because as you are collecting the data it could be all kinds of reasons. I wonder aloud what will be commoditized first. The processing capability on some of these platforms or the data itself. I think that you are right data is a new asset for the new regime. It is a new paradigm to think about data as a key and central resource versus as you mentioned an appendage that sort of doesnt feel particularly critical to the operation of our programs. We do all kinds of domestic policy military policy around who can access Rare Earth Minerals<\/a> and various forms of energy. And the nationstates access to the data set that has been properly stored and collected are viewed just as precious. I also wanted to reflect on the smartest hour that i ever spent listening with our chair and Ranking Member<\/a> tod discuss some ofof these issues. What was presented in the argument is that china understands ai control of governance is equally a threat to them in the United States<\/a> and so thet argument was we are idl partnersse with china because we share a common goal to not have the robots take over the government. And the chairman offered i think a pretty stride critique of that program saying while we view they are taking longterm and shortterm theyt are more genocide than Team Humanity<\/a>. So i wonder because you have so much in the written testimony about your time in china and how that shaped your perspective on this, do you think that china sees the interest with of the United States<\/a> on this or do they see us as and armslength competitor . I think that it would be a stretch to say we are on the same team on this issue. I think if you look at the last generation of the Computer Vision<\/a> technology, the way that china approached it was utilizing building the Industrial Base<\/a> that was governmentfunded to immediately build advanced technology for these a person of the population. Ultimately tightening the grip of the totalitarian regime. I expect them to use modern Ai Technologies<\/a> in the same way to the degree they can and that seems to be the immediate priority of the Chinese Communist<\/a> party when it comes to implementing ai. Just a question for the record i wouldld love to know everyones perspective on what the most Important Alliance<\/a> is when it comes to these regimes. Is it august, does nato have a role to play in some of the ethics around this, so i would like to submit that. It is a corollary if you Say Something<\/a> nice about me or the Ranking Member<\/a> you get more time. Just quicklyme but is the answe . It is an interesting question. Within that group as we look towards Artificial Intelligence<\/a> as a Global Technology<\/a> that will shape much of the future of the world i think that we need to form and as many key partnerships as possiblele to ensure certainly for military use and use in intelligence and for use serving commercial purposes are adhering to the Democratic Values<\/a> that we have as a country. From a traditional Security Alliance<\/a> ie would say they will be critical however i would say that the broader Economic Partnership<\/a> with our friends and alliesnd and the European Unions<\/a> going to be critical longterm and is going in the wrong direction. Happy to talk about that more. I echoed the sentiment of the others. Later in the year we will have our first summit that is happening in the uk. So we need to expand the Cyber Alliance<\/a> as mentioned earlier on the area of ai. Ir would say following on tht last question. We have had to generations of learning how to share with each other and become interoperable. I dont know if we have data sharing arrangements when we dont have a joint platform. Its fascinating to think about it like getting those arrangements that took place and sharing the data given the value is going so precipitously up on it. With the help of industry thats invested in ai is like admiring the problem. Weri are all talking about the problem of this new tool that we know has real potential but also has potential downsides into so how do we govern it and our constituents are asking us what are the ground rules on this new technology because it sounds scary. And i would come into the joint Artificial Intelligence<\/a> center grant dod for putting up some basic 40,000foot guidelines on being responsible and equitable and traceableta and reliable, bt its like its real top level staff. But we are up here the flip phone generation trying to figure out how to govern ai and its complicated. But could you give a sense sort of in colloquial english off what keeps you up at night about the military use of ai . If china is investing at least three times as much in some cases ten times as much as we are, what is the number one thing that you feel like kind of worst Case Scenario<\/a> if we go unchecked we could see in the next decade . You are shaking your head. Thank you. While there are certain risks of what we would call the spoke threats, i think the most acute challenge we are likely to encounter in the near term is simply more effective and efficient enemy. So the chairman referenced a quote and i will raise with another one from general pershing who said infantry winds battles but logistics wins the war and the supply chain and military logistics and a lot of what we call back office military capacity which can sound not so scary. Mode after watching russia and ukraine. Ib would be happy to improve the logistics given what weve seen in the buffoonery end of the Russian Military<\/a> but i just want to sure that mr. Wang has an opportunity. And its not a scary thing. Its just a more capable and competent adversary whoever they are. I certainly agree that to back off the function logistics oand of the optimization is critical. If you look towards the areas where they are investing in the Artificial Intelligence<\/a> it is for the autonomous drone forms and whether that be aerial, subsurface or ground investing across all fronts. They are investing into the adoptive Radar Systems<\/a> that jam and blind the sensors and information so they are investingne across the whole spectrum and Artificial Intelligence<\/a> to sort of set the tone of warfare with this technology so we need to be investing across the slate. I worry about the risk deploying the systems without proper guardrails and for me it comes down to implementation which is Test Evaluation<\/a> so how do we know that for all of the Artificial Intelligence<\/a> systems that the dod is likely to deploy over the next few years and decades and how do we make sure in the year or two it is as stated to so it needs to be a standard part of the procurement process as the test and evaluation mechanism to ensure that every instance where a Program Within<\/a> the dod is looking to the intelligence and do we have the right testing and evaluational to ensure that it adheres to the guardrail. I know the department is working hard on the data labeling problem. It tends to be stovepipe the organization to share the data and make it available, make it usable. If you could get them to do one thing on reliability and availability, what would it be . I would say first establishing the Data Repository<\/a> and then creating a plan by which as much out of the terabytes go into the central Data Repository<\/a> and then creating a plan by which as much of the data is processed and labeled and annotated to be ai ready as possible. These are multiyear efforts that are not going to be solved tomorrow with the snap of a finger. They need too be solved in longterm planning and coordination. Thank you. I yield back. Thank you, chairman and a chairman gallagher for your leadership and to the witnesses for helping to inform congress on these important issues along with my colleagues from both sides of the il im concerned with of the advances in our Artificial Intelligence<\/a> and Machine Learning<\/a> specifically with adversaries like china. What concerns me most is the communist party making Great Strides<\/a> and intends to be the leader by 2030. While we here in the United States<\/a> made improvements in recent yearsye we continue to advance we still have much work to do when it comes to entering the dod is adopting and deploying the capabilities properly. To give a plugt for a piece of legislation. Myat legislation will require te office of management and budget guidance to federal agencies relating to the use of ai and Machine Learning<\/a>. Specifically when being used to supplant a humans decisionmaking impacting the citizens. While focusing more broadly i want to ask for your thoughts on whatou the dod currently stands with when it comes to the ai and Machine Learning<\/a>. Where is the u. S. Compared to the adversaries such as russia and china with implementing the latest capabilities over the years ahead, on par, are we years behind and in some ways what we planned for them to speed up the adoption and implementation of ai effectively at the department . So, when we look at the new Technology Like<\/a> the language models. We need to implement as possible and we will see how that develops. Within china they quickly raced ahead so theyve built an Industrial Base<\/a> within the next country and funded with government money to build facial Recognition Technology<\/a> which should be deployed through the country to suppress and overall tighten the grip of their socialist regime. If you look today at the leader boards for the ai competition globally, Chinese Companies<\/a>, chinese universities dominate compared to the american institutions. So if you look at that as a case study, the chinese clearly have an ability and well to race forward when it comes to the Artificial Intelligence<\/a> deployment. 2 terabytes of data every single day so we can properly build an advantage orwe it will be quite durable. Mr. Kitchen which you add something to that question requires your scorecard at . Are we behind, but head or on par . Works the two global powers were competition matters most is between United States<\/a> and china. As i mentioned at the beginning of my testimony, a year ago i very real concerns as to how the United States<\/a> is going to be able to maintain its ai s advantage. But, precisely because so much of the m conversation on ai legitimately so, the public conversation focuses on the risk and the kind of unknown. Again meaningful conversations. I do think analytically i do believe that we have a moment to reassert american dominance in a way that really matters but some sothings i would have called a drag on our development and deployment from National Security<\/a> perspective are actually lessening. If we realize this technology deliberately, then we can seize the advantage and how to seize the advantage now but actually built in advantage that will thbeat meaningful over the longterm. Think we should do everything we can to do that, thanks regards 30 seconds ago doctor alessi last question, were the risks that this committee and the department should be aware of . How should we address those risks as we move forward . Looks so at risk theres obviously a fallback if the United States<\/a> fallsls behind wih regards to the ai military. The main area we need to focus on is to makere sure that we do have the advantage and research investing in the research pretty specially in the military side and making sure we are still a leader in the area of rnd and aip. Thanks auteur my time is expired. Think it was her chair. Thank you so much for coming on out and talked with us today. We spent a lot of time today talking so far about whos got the Development Edge<\/a> and it will be a building that direction fot certain about competition with china. I do not want to go over those right now. As they were very well talk to through. Mr. When you talk to your Opening Statement<\/a> some of the Main Technology<\/a> of the past being about Nuclear Development<\/a> and whatnot. We are shaping this era in this very well likely shaping our era. I want to get a sense about what you think proliferation of this technology impossible weaponry would look like . We are in the nuclear era, which we still are. We have a situation here were only a very few set of countries have been able to reach that threshold of technology and proliferation has been in many ways tried to be contained in that capacity. I guess i want to ask you, for me that is not necessarily see make the kind of set up we are likely to see over the coming decade or two. What does it look like to you . Are we going to have a situation you ask and china and a handful of companies are Major Developers<\/a> and gatekeepers to this technology but the actual Weapon System<\/a> and technology will be potentially mass deployed and able for purchase by pretty much any nation that is out there . Give us a sense ofiv what that topography and landscape looks like because thats a really good question. I think in terms of impact, Artificial Intelligence<\/a> will be similar to nuclear weaponry. It is a Technology Like<\/a>ly to be ubiquitous. A, Artificial Intelligence<\/a> can be usedin across every single gentlemen, every singlesi function. Every single activity at thehe e military hasy today. Its not contained as one individual weapon. It is the technology that is increasingly becoming a Global Technology<\/a>. A few months ago, the uae announced their own they had built called falcon, they open source that model to the world. Anybody on the internet can go and download that model the open Source Community<\/a> when it comes to Large Language Models<\/a> he technology is likely to be accessible in some way, shape,to form to nearly everyone in the world. That being said, i think thats not a reason to give up hope. Because one of things i mentioned before on military applications, you need trained on military data. Mr. Kitchen id like to bring it in when we find a situation that yes, some country or entity or company is doing that but then able to then sell that type of technology and a weaponry to a country or to a group. Id like to also get your thoughts on potential for rogue actors, nonstate actors to get the type of technology to utilize it. If you dont mind give us some your thoughts. I agree with alex in the sense this technology had the same strategic impact of some Something Like<\/a> nuclear weapons. One of the peculiarities of it is this technology is overwhelmingly being developed in the private sector unlike nukes. And so one of the implications of that is that so much is done via the open source model. Its instant proliferation. It is available in terms of the underlying capacity. Its going to be the applications the particular applications that really make the difference when it comes to capability distinctions. That is where alex a point about potential advantage on military data. How we apply its really, really going to matter and thats with the advantage comes to us. We think about nonstate actors or rogue actors, i think of awarethe most acute challenge te is probably on novel in traditional cyber exploitation of these capabilities. An automated and deploy it is now going to be democratized to a level in scale. One last question, doctor to bring in on this we talk about the proliferation seeing the potential for nonstate actors and others we talk about some these frameworks the most easily the weights should we be thinking about it at National International<\/a> agreement here an International Treaty<\/a> . What structure should we be building towards get our ability to structure that as a whole . I completely agree we need to think about the domestic side. Here in the night states me too think about how we should be governing these type of technologies, understanding the risk and having a mitigation process in place. But we do need todo work with allies S International<\/a> level. Okay thank you, youll back. Mr. Valdis reckon uss for fie minutes per. Think i one to follow bro quickly. Ransom ware is a huge problem already. Largely because under the radare unless Colonial Pipeline<\/a> is hit or gbs and thats what everybody talks about for a week and then forgot about it and then acted as ifd its not a real problem. Which it is your friends and industry in small companies, 102 hunter people getting hit. Half a Million Dollar<\/a> ransoms are being asked in a a million Million Dollar<\/a> ransom is 50 grand a few years back. Do you think with ai were going to face as you just mentioned way to expand on it than explosion and or transit where we talk about democratized . That is certainly one of the potential implications. I think the Key Developments<\/a> over the last two years has constrained ransom or to the degree it has been constrained ukraine. Ere on many of those cyber syndicates that were prosecuting those attacks have been repurposed by the russiann government for attacks in ukraine and elsewhere. I think if and when that ever slows down we are going to feel the surge again. I think that surgical absolute be enabled by generative ai. One of the key areas, theres a study that says four key areas that will constitute a proxy 75 of the economic increase comic ngwith jen ai part one is r and. Software development being the other. And so i think that applies equally to bad guys as it does the goodal guys. Nobody is ever accuse the dod of being highly efficient. They are large. When you have inefficient seizures on what wasting billions of taxpayer dollars. Particularly when we are in a competition with china thats even more troubling and we need to address it. We might envision ai with future wars being fought by robots and such. Within the walls it self, these wwells, mr. Wang in your opinin can the department of defense use ai to extract efficiencies deficienciesin programming and y activities . Were sure part one of the areas we have already worked with some ofth our duty customes on is using Artificial Intelligence<\/a> and a large link models to help digest requirements that are given by the dod. There are so many groups with the department of defense are generateden requirements and a matching those requirements up with capabilities in the private sector or dod develops is an incredible efficiency potential efficiency gain. There are hundreds if not thousands of applications like that of Artificial Intelligence<\/a>i nc dod a moreg the efficient organization. So im incredibly optimistic about the ability to use ai it whether its logistics, back office, and personal relatedu matters to build a more efficient force waste your resources ultimately is able to have more force a production capability per. The same thing holds for increased accountability duty contracting and spending . Quick to think about the limitations are with the challenges are its in processing huge amounts of information that is being generated by the dod to understand not only how funds are being used but understand the capabilities are being generated are. Ifif you think about the probles that its naturally suited for Artificial Intelligence<\/a> and for the use of the large models. Dr. Kagan you talk about ai my mind starts to bend and hurt and or break a little bit because it is so intriguing. We talk about basic concepts of some of the technology were not accustomed to like social media but some folks believe it or not on the other side of the building do not grasp even those. The night states governors and we should use twitter war did not get the name right. One of the senators sing like how can they post a picture on the line . Things like that. Well its funny its also troubling. Theyrey not grasping a basic concepts your jacket ai with this stuff on it. Our colleagues in the American Public<\/a> on ai and dissuade some of the fears . So many are thinking about the education side of it. We need to understand this education is to be tailored toward peoples needs. So, depending on their role, depending upon their responsibilities need to tailor that education for them. To get them an example for leaders who may not be technical we need to come up with an education listen to it ai is exactly to your point but what is capable, what its limitations are. Versus someone who is, technical for them that will be different story. Have a more Technical Education<\/a> for them. But also having education in a continual form o of ai involved. Almost like how it can help them specifically and make it riseth a little bit better becae her. Exactly. Think it was determined i yield back. Recognizer five minutes for a quick thank you mr. Chair, good morning thank you all for being here and for your insights. I want to build on written testimony. You talk about data is the ammunition and ai warfare you talk about particulate china are doing. Based on your specific expense and your companies are working with dod. Who is doing relatively better, one of the lessons we can learn question rick also if you can talk about cda oh and how you see that intersectingg here so that we can recognize the imperative and wrap your arms around the data better. Certainly. The groups we work with by us generally quirky with the more forward leaning groups of the dod are forwardlooking prediction innovative in terms of taking on this technology thats a key part of the go Forward Strategy<\/a> and building capabilities. We work with many of the early programs and the dod for use of ai. And by and large i have been incredibly impressed. That being said now isng an opportunity for us to build on the successes and take this olmoment and speed up the moment its incredibly important to build our past successes. We are able to more scalable he deploy the technology across entire dod rather than being limited to a few innovative cells within the dod. As i mentioned a bit ago they di you are in cao have been some of the groups the dod. Weve been able to have fast procurement cycles in general innovate when it comes Artificial Intelligence<\/a>. That is to happen across the entire department of defense. Lastly, on they have done a recently established organization theyve done a good job of pushing forward and building the labeling in the depository for the dod. Now we need to ensure that actually happens in terms of collecting the 22 terabytes of data that are being generated every day per. Thank you and to build on that a break and anyone else who wants toto adhere, is it even possible to do that from the top down . Ith understand the importance of setting the right tone and direction. If we think, creating a new office is necessary i would argue not sufficient if we are serious about wrapping our arms around this. It should be emphasized, trained and reinforced much more broadly, do you agree with that question rick any ideas on any but how to do that . Particulate looking how others emissaries and allies are doing it . Is necessary here. The individuals making the decisions when they get a new hard drive off of the military platform they need to make the decision what theyre going to the hard drive we needke all the doubts that individual to understand that hard drive is full of data that will fuel the future of the military leadership we need to understand that as this really with you for a topdown perspective within thispe conversation. It requires a whole dod approach to properly achieve the outcome. Congress on the one thing i would add is as we tackle these difficult challenges and they are legion, from eight mentality standpoint i would encourage congress the u. S. Government to approach these challenges that have to be managed not solved. If we tried to find the exquisite solution we will so delay ourselves we will miss the opportunity pray that is one of the key narratives im really trying to emphasize that we really do have a meaningful strategic opportunity. These guardrails everything matter, they really do. As we approach these things, seizing the opportunity is probably one. And that doing well and carefully as part of that. It cannot be the goal by which we have to leap over. I appreciate and agree. Very briefly he hit on all little bit your sons and the researcher rome academic aroma how are we doing there . What can we do better . I think i could guess. We can definitely when you are and investing in the Research Side<\/a> of this, it opens the door for us on the innovation side to also invest in research on the safety aspect read the guardrails that were mentioned. When you are investing in the research and order to were always ahead. Thank you and i yield back requested a second round of questions, all we can buy recognize myself for five minutes. I want to return to mr. Gates question about key allies in the alley competition. You all mention our most obvious ally. I am not detracting from that answer. I like to invoke the concept of geopolitical swing states. Perhaps countries that may not fit neatly within the free world paradigm. What are the emerging ai superpowers we may not be thinking about . Or less you sayki states that punch above their weight when it comes to ai that we need to be cultivating and ensuring theyre not in the party direction. Well start with you mr. Wang. I do think its very important as ai promises to be one of the most important technologies both economically and militarily there are a myriad of countries getting involved. That uae has a very dedicated effort towards Artificial Intelligence<\/a>. Open source models they are continuing that series of develop its towards building they have more powerful ai models we do not they would open source and we will see. As theyn develop and do it we cn to make sure they follow our principles in our governance regimes. India is another key country obviously very critical we think about geopolitical allies. But also as you think about their development in ai there incredibly they have stated efforts to develop large language within their country. We want to ensure our thinking about it Artificial Intelligence<\/a> is in fact the same way we ares country. I would agree completely. This affects the way we think about our relationships. Right now our Technology Supply<\/a> chain is distributed in such a way as to where critical vulnerabilities many of the keynotes are deep within chinese fear of influence. Or i think we are going to be going as we are going to try to build trusted Technology Ecosystems<\/a> amongst trusted partners and allies for the idea is we identify western democracies as being the type of organizations that we can partner with so have mutually beneficial trade and Technology Relationships<\/a> of future partnerships. That requires a common purpose and a common understanding of the opportunities and challenges. When things im most concerned about is where many of her friends and allies are in the European Unions<\/a> on this issue. My point there being that when we think of military and operability of these alliances we also need to understand military is going to be predicated on a regulatory. That is where we have a real gap between us and some of our key friends. The European Union<\/a> seems to have concluded to build their own domestic technologybased they have to deliberately constrain and at times even decouple the american technologybased. That will not work for a shared purposes it will be a real problem Going Forward<\/a>. So i completely agree with the weaknesses in regards with alliance. One of things we need to understand also for those type that were mentioned we need to think about how we can align ourselves with them to make sure their advancement and ai is aligned to the ignited states. So we would have that alliance while we are ensuring they are still the leader in this space. Mr. Kitchen, you mentioned in written and oral testimony on hardware we have a strong bipartisan consensus allowing us to constrain chinas advancement. There has been some suggestion as we game with Silicon Valley<\/a> leaders while we admire the gpu export controls i get the bite of ministration credit for tha. There are loopholesed whereby ty are still able to access the second most advanced chips now. I recognize i run out of time here. This goes to my previous point about interactive process. The export control that was the first and we are beginning to optimize and tighten those controls. Some surprise government andno industry are doing back and forth on this. Between both stakeholders the actionis necessary now are tryio find the right way for it i have high confidence where it will do that. Mr. Wang questioner. It is true you can see reports other Chinese Companies<\/a> have bought billions of dollars of gp used in the past. It is in this year so far. It is something that we need to be externally careful and vigilantly about. When chairman gallagher and i had that conversation with elana mosk he said agi was five six years away. I was surprised by that timeline. What is your sense of how long we are from agi . Looks agi is an illdefined concept. Could you define it so not doing acronyms . [laughter] you are the guy, sorry. [laughter] agi stands for artificial general intelligence for the idea we build in ai that is generally intelligent in the way humans are. It is not a super well defined concepts. Even and he is in the current ai systems youll notice clear limitation and issues and challenges they have with them doingeven things like basic mat. Agi as a concept is an enticing one we can Silicon Valley<\/a> talk about a lot. I do not think its a very wells welldefined and certainly not something that should immediately affect how we think about putting 1 foot in front of the other. Not only Economic Leadership<\/a> as well as militaryel leadership. The technologies today large instant models inc. Developed and deployed today from an Economic Perspective<\/a> or from a military perspective. Thats why think its important to set the foundation to them investing in data. Set up the foundations for longterm success last concept is it comes to a g. I. Production timelines at often a way to distract which is in my mind very important. I think alex is exactly right. The idea of Artificial Intelligence<\/a>, increasingly agile and Capable Foundation<\/a> models of the ai capabilities they can be more broadly cap will. One of the features of the foundational models as emergent capabilities. It is the idea that we created this Foundation Model<\/a> to do a particular task and lo and behold it actually can do this other thing without having been trained to do so. So we are going to see that that is a common feature. I would say the timeline that was given to you about Artificial Intelligence<\/a> in the next five years is aspirational. If it is good. Looks similar to the previous comments, it is aspirational. But, what i would add to that is we are headed that direction. In regards to Foundation Models<\/a> these type of models that can provide tasks that they are not pleasantly trained on but they can generalize it. However, while we are heading into that direction, obviously not in five years but we need to all but we are investing on the Research Side<\/a> we also need to invest into the safety aspects of it to make sure youre able to mitigate the risks you are anticipating with regards to artificial generated intelligence. Quickly asked my last question which is do you think we need any dod clearance for any types of ai like we have for Nuclear Technology<\/a> . There are safeguards only so many people to get access to it. Mr. Wang is there anything in the ai space . If we think toward military ai systems so much of the next generation of capabilities are going need to be built and trained on already classifiedlr data theres art existing structure regime to protect any models or trained on classified data whether its topsecret or beyond level to ensure those capabilities stay limited to certain state control. I dont know if we need to build more on that but i think it is certainly true most of the capabilities the dod looks to build are likely to be developed at the topsecret level. Thank you. I am interested integration on ai in Human Performance<\/a>. We always a very touched when ever we have casualties that are in training or otherwise that are preventable. What have any of you learned about were some of the potential lies and utilizing ai and integration and Sensor Technology<\/a> or other Human Performance<\/a> capabilities . One of the areas where Artificial Intelligence<\/a> i think has the greatest problem is doctor kitchen mentioned before if you look at one of theof largest costs of casualties in transporting fuel and other resources for the military for this is an area where autonomos vehicles or even heater followers set ups are able to greatly improve efficiency as well as reduce casualties for the military is one of the goals of the army robot robot combat vehicle we are collaborating on. As we look further ai systems are Assistive Technology<\/a> in with our scale platform we are able to assist in key decisionmaking with his being utilized right now it in military planning exercise to help ensure all the data and information the dod has access to his being integrated into the correct military decisions. There is an incredibly Bright Future<\/a> i think for assisted use of Artificial Intelligence<\/a> to make the dod more effective. Some of the most excitingab things about ai in my view is its ability to help expand human thriving. Many will have seen a commercial with one Technology Provider<\/a> who uses their phone could help speech pathologies or difficulties communicate more effectively. The open ai the check gpd has a tfunction for vision impaired individuals and describes images for them so they can participate and knowledge gained an application we think about military context with going to be ai Underlying Technology<\/a> and capabilities of the naples opening from allowing again tocs to walk brain injury prevention and recovery. The things this technology again im not an idealist on this but the promise is real. I what it means for our society just in general is very promising. Is sought of a paraplegic mother that its an inspiring ig concept her doctor, i went to paschal different twist on the questions you have talked to my colleague to some degree about how we ought to measure the soft power capabilities of some of the ai platforms. How is it our thinking about what it would mean to the United States<\/a> as opposed to china to be a leader in deploying one third thousand ai robot doctors into africa or latin america or or somewhere else on the globe . It is all about howll we want to have our values embedded into the ai systems. Thinking about the principles, one area special the dod has his assistant to governable. Depending upon the level of risk the systems posts we want to have oversight. In some cases the risk is low in they want to let ai make the decision. Imagine a benign example recommending a movie. But in specific cases especially ones thatse are lethal we do not what ai to make the decision we would human oversight. We went ai to be used to provide us information. Patterns we may not have seen. We would use that information and us humans would be able to make the judgment. These are elements we to consider. Still substantially impact scalability and being able to deploy the check i would t thin. So we have a company and some government processes its not necessarily reviewed as an obstacle with capability that has a governance process enables us to have standards, policies in place i can easily apply to any ai case that we have. With that foundation of Ai Governance<\/a> we would be able to replicate the process for any ai usage. Thank you have not given enough time to answer the question mr. Wind that where the things im sure would like to explores further when we get into the test and evaluation paradigm that you keep coming back to in your testimony its important for us to get a comp corporation principles of what that test and resume look like i hope youll continue to work the subcommittee on that. I yieldrk back. Victims are going to do a third does one go very quick trust me. Going to apply what i call the references might 96yearold grandmother very smart even a member of the flip phone and generation little in the ai generation. So want you to imagine youre sitting across from my grandma, you each have an old fashion is an herb bait has him as a world war ii vet. You need to explain to her what she needs to know about ai, why this conversation matters both for the future of warfare as well as her life in the lives of her children and grandchildren. What do you say to the great in beautiful virginia justice . Actively look towards world war ii in the last air of conflict, new technologies like the atomic bomb were critical in ensuring we both had American Leadership<\/a> in that the values america upholds work able to continue to prosper and set the tone for its development to the world. We are now embarking on a new era of the world, one in which a new technology, Artificial Intelligence<\/a> is likely to set the stage for the future of ideologies, the balance of power the future of a relative piece of our world. RlArtificial Intelligence<\/a> is incredibly powerful technology that underpins nearly everything that wee do from an economic and military standpoint and therefore its critical that we as a nation think about how we not only protect our citizens from risk of Artificial Intelligence<\/a> but also protect our ideologies and democracy to make sure we continue to be leaders. Mr. Kitchen . Maam, there is a new technology that under the right circumstances could put your grandchildren and this nation. That could make thiss nation economically and militarily Strong Enough<\/a> to defend its people and its interest. In the technology that in the wrong hands could impel the same things its really important that your government and industry Work Together<\/a> to realize those promises and mitigate this threat. Okay, doctor brickley it is a technology that is pretty much embedded in our daytoday life. They are living with that there breathing with it. We want to make sure it is technology that is part of our life. It has our values the values we foughtht for. It is incorporated into the basic technology so we would still have our Civil Liberties<\/a> and civil rights as well as using this technology and levering it to have a better quality of life appeared. Okay, bye though it just occurred to me though i loved gallagher if i had my mothers maiden name justice i would probably be president at this point per. And a progressive. [laughter] well played, any other questions . Kirks okay a bit of housekeeping before we adjourn. I want to enter three things into the record quickly the first is the article referenced bbefore bite jericho and on geopolitical derisive geopolitical swing states published on may 15, 2023 the second is something you mr. Wang wrote in november of last year on the ai war and how to win it and what you say we must recognize our current operative model will continue under trajectory for the next 10 years result in us falling irrevocably far behind. White at Large Organizations<\/a> continue on the path to the demised even if they future is painfully obvious thee reason s inertia bureaucracies will continue to provide deep into the abyss for an attorney the third is a recent article by Mark Andreasen<\/a> optimistic case for ai and thought why ai will save the world in which he says single greatest risk of ais china wins global ai dominance in a week, the nine states the west do not. I propose a simple strategy for we do about this back the same strategy president Ronald Reagan<\/a> used when the first cold war the soviet union. Which is we win and they lose. I ask unanimous consent to enter all three of those into the record. Without objection so ordered. Ask unanimous consent never set five days to submit statements for the record. In the hearing stands adjourned. My [background noises] [background noises] [inaudible] [inaudible conversations] [inaudible conversations] tuesday a hearing on how to regulate Artificial Intelligence<\/a> Technology Senate<\/a> judiciary subcommittee hearing is live at 3 00 p. M. Eastern on cspan three. Cspan now are free mobile video app or online at cspan. Org. Sees fans at washington journal a life of form involving you to discuss latest issues in government, politics and public policy. From washington d. C. And across the country, coming up Tuesday Morning<\/a> dental Services Reporter<\/a> Brendan Peterson<\/a> of punch bowl news previous this weeks federal open Market Committee<\/a> meeting and another potential Interest Rate<\/a> hike. We will talk with adam founder of open the books. Com about his research into waste and abuse in government spending. Also a discussion on Climate Change<\/a> policies recent extreme weather patterns with executive director of a public bob and david president of the Environmental Defense<\/a> action fund. Cspan at washington journal joins in the conversation live at 7 00 p. M. Eastern on cspans he spent now or online at cspan. Org. Nonfiction book lovers, cspan has a number of podcasts for you. Listen to bestselling nonfiction authors and influential influence on the after words podcast. And a q a here wideranging conversations with nonfiction authors and others who are making things happen book notes plus episodes are weekly hourlong conversations that regularly feature fastening authors of books on a wide variety of topics. The about books podcast takes you behind the scene of the Nonfiction Book<\/a> Publishing Industry<\/a> with insider interviews, industry updates and bestseller list. Find all of our podcast by downloading the free cspan now cap. Or whatever you get your podcast and on our website cspan. Org podcast. Cspan is your unfiltered view of government but we are funded by these Television Companies<\/a> and more including while. The world has changed today the fast reliable Internet Connection<\/a> is something no one can live without. So well is there for our customers with speed, reliability, value and choice. Now more than ever it all starts with great intranets. Wow. Whe support cspan as a Public Service<\/a> alo with these other television providers. Giving you a front row seat to democracy","publisher":{"@type":"Organization","name":"archive.org","logo":{"@type":"ImageObject","width":"800","height":"600","url":"\/\/ia802701.us.archive.org\/4\/items\/CSPAN2_20230725_060100_Hearing_on_Threats_to_Electric_Infrastructure\/CSPAN2_20230725_060100_Hearing_on_Threats_to_Electric_Infrastructure.thumbs\/CSPAN2_20230725_060100_Hearing_on_Threats_to_Electric_Infrastructure_000001.jpg"}},"autauthor":{"@type":"Organization"},"author":{"sameAs":"archive.org","name":"archive.org"}}],"coverageEndTime":"20240707T12:35:10+00:00"}