Boost Your Marketing Results with Cookiebot’s Powerful Cookie-Based Solutions

123456789101112131415161718192021222324252627282930313233/ 33100%实际尺寸适合宽度适合高度适合页面自动介绍杂志连续的

Delivering on artificial intelligence’s potential Insights from • Army • Brookings Institution • CISA • DoD • GSA • National Geospatial-Intelligence Agency • National Reconnaissance Office • National Technical Information Service • Office of Science and Technology Policy • Upturn EXPERT EDITION BROUGHT TO YOU BY

TABLE OF CONTENTS Investing in federal AI skills …………………………. 4 Achieving data readiness for AI/ML …….. 7 Accelerating federal AI adoption …………………… 10 Driving AI to the tactical edge …………………. 14 Securing federal supply chains ………………. 17 Implementing AI identity proofing …………. 20 Streamlining Army selection board processes ……………………… 23 Embracing the art of the possible ………………….. 25 Expanding IC partnerships with industry ……………………….. 28 Moving up the AI maturity curve …………….. 31 Can AI let your teams use their brains better? That headline is not meant to be demeaning. It’s a serious question, and one for which the answer appears to be a resounding “yes” — from both federal technology leaders and industry developers of artificial intelligence and machine learning technologies. Most people agree that one of AI/ML’s greatest potential benefits in government is that it can free data scientists and subject matter experts from the tyranny of clerical and mundane tasks. They then can use their expertise to wrestle with mission challenges and other complex demands. There are more than a few examples. Here are two: • Consider federal supply chain risk management efforts. “There really is a very shallow pool of subject matter experts out there in this area. Because that pool is so shallow, we have to turn to automation to help us,” shares Brian Paap, cyber supply chain risk management lead at the Cybersecurity and Infrastructure Security Agency. • Think about decision-making on the battlefield. “At the edge, the ability for things to be deployed and automated versus needing large teams of people to come in and to do those deployments is going to be critical. Warfighters are very talented folks, but they may lack the IT talent at the edge to do this. And that’s where we think automation can be helpful,” suggests Jim Keenan, vice president for DoD at Red Hat. In this ebook, we share strategies and tactics for accelerating and maturing federal AI/ML initiatives, along with details about implementing new technology tools, establishing appropriate guardrails and developing metrics for success. In the 10 articles, you will discover advice and insights from multiple agencies as well as AI leaders in industry. We hope it will help your organization mature its own use of AI/ML as you strive to make smart decisions faster by relying on data. Vanessa Roberts Editor, Custom Content Federal News Network FEDERAL NEWS NETWORK EXPERT EDITION: DELIVERING ON ARTIFICIAL INTELLIGENCE’S POTENTIAL 3

With the National Artificial Intelligence Initiative Act hitting its two-year anniversary, federal leaders are looking at more ways to invest in their workforces to better implement AI tools. Although understanding the talent and skills needed to better take advantage of AI is key, there are additional barriers to implementation. “What we have is a large number of federal agencies that are struggling with antiquated architectures and a lack of skills and talent,” said Chakib Chraibi, chief data scientist at the National Technical Information Service at an ATARC event on implementing AI. AI can play a crucial role for federal agencies, if they are able to implement it effectively. That means creating responsible guardrails like privacy, transparency and fairness in the use of AI, Chraibi said. At NTIS, “it’s helped us make evidence-based decisions, improve on customer experience, perform intelligent automation, enhance data privacy and ethical data practices, as well as strengthen our cybersecurity systems,” he said. Some agencies are already aiming to make more internal investments to boost their workforces’ understanding of AI, as well as train current employees on best practices. The Army, for instance, is looking to do more internal upskilling and recruiting, while also working to maintain industry partnerships. “It’s key for anybody embarking on an AI journey to know and understand your organization’s mission and how AI can enable it,” said Army Forces Command Chief Data Officer Jock Investing in workforce AI capabilities Agencies should make internal workforce investments to improve AI implementation, experts say BY DREW FRIEDMAN 4 FEDERAL NEWS NETWORK EXPERT EDITION: DELIVERING ON ARTIFICIAL INTELLIGENCE’S POTENTIAL

Padgett at the ATARC event. “You don’t always want to outsource your data and AI talent, so invest in your people upfront.” The Army also wants to add to the internal team that deals directly with AI-related work, Padgett said. Although the Army can make direct hires for software developers and data engineers, the service still relies heavily on the private sector to hire data scientists. “Data scientist talent is very weighted on the industry side right now. What I do see happening over the course of several years is that scale will end up balancing itself out to some degree, as DoD as a whole starts taking on the training tasks, new skill sets [and] upskilling,” Padgett said. For the Defense Department overall, Jaret Riddick, DoD’s acting principal director for trusted AI and autonomy, said that diversity, equity, inclusion and accessibility also play a role in the recruitment process. The Navy, for example, recently invested roughly $27 million to expand its Historically Black Colleges and Universities/Minority Institutions Program, Riddick said. These types of investments help DoD with “expanding the aperture to look for talent.” “Down the road, there will be a critical need to grow the talent base and to maintain an eye on the capacity of the industrial base in the future, to produce these technologies that we’ll need,” Riddick said at the ATARC event. AI is not the only area where DoD is looking to expand its connections with HBCUs. In June 2022, DoD and the Air Force created and funded a new research institute, partnering with 11 minority institutions to create the organization. Along with these types of minority institution partnerships, DoD is adding other industry partnerships as well. “We are promoting the growth of new companies, startups and small businesses [and] we are, of course, engaging with the traditional players,” Riddick said. To best implement and use AI, at least some understanding of the technology is necessary at all levels of an agency’s workforce, Chraibi said. By assessing the internal resources and skills Data scientist talent is very weighted on the industry side right now. What I do see happening over the course of several years is that scale will end up balancing itself out to some degree, as DoD as a whole starts taking on the training tasks. — Jock Padgett, Chief Data Officer, Army Forces Command DoD expands investments in HBCU programs FEDERAL NEWS NETWORK EXPERT EDITION: DELIVERING ON ARTIFICIAL INTELLIGENCE’S POTENTIAL 5

that are currently available, agency leaders can then focus on upskilling and training where it’s needed. They can also identify what support they may still need to obtain from external sources. “It’s important to have leadership understand this technology, understand what are the needs, what are the requirements, and of course, supply the skills and resources that are needed to be successful,” Chraibi said. It’s important to have leadership understand this technology, understand what are the needs, what are the requirements, and of course, supply the skills and resources that are needed to be successful. — Chakib Chraibi, Chief Data Scientist, NTIS 6 FEDERAL NEWS NETWORK EXPERT EDITION: DELIVERING ON ARTIFICIAL INTELLIGENCE’S POTENTIAL

How agencies can achieve data readiness for AI Agencies across the federal government are embracing artificial intelligence and machine learning, but the first challenge they often run into is preparing their data. Data readiness involves having a managed concept of data that allows the data to understand what model of AI is being applied: machine-machine, machine-human or human-machine. Having data readied will lead to greater speed, agility and transparency in the data. Enrichment of metadata is a good way to accomplish this. It lets the machine understand more about the data, whether it receives it from a human or another machine, and it allows humans to know more about data output by a machine. That can answer a lot of analytical needs for additional information about the data, as well as supporting the query layer of AI systems. One effective way to do this is to store the metadata alongside the data itself. “Not only do you get traceability with the data being alongside your metadata, but you also get that information in a very fast and effective way,” said Bill Washburn, chief program officer at MarkLogic Federal. “If you’re doing a geospatial search and you’re looking for a plot on a map, or if you’re looking for a section of information through a visual acuity — like not just a map but maybe a video, that information being stored alongside means that I only have to search for that information,” he added. “There is a great advantage in having those two things be symbiotic.” That’s especially helpful when dealing with unstructured data, which the government deals with in droves. The Defense Department and law enforcement agencies are applying AI to video analytics. The intelligence community, Interior Department and agencies like the National Oceanographic and Atmospheric Administration make frequent use of maps and satellite images. And most agencies, including the Veterans Affairs Department and the National Archives and Records Administration, are working hard to digitize paper records. “An advantage of NoSQL is you don’t have to convert images to text. You don’t have to change Managing unstructured data An advantage of NoSQL is you don’t have to convert images to text. You don’t have to change it. You don’t have to wait for it to be modeled. — Bill Washburn, Chief Program Officer, MarkLogic Federal PROVIDED BY MARKLOGIC 7

it. You don’t have to wait for it to be modeled,” Washburn said. “That can be modeled as your data is, rather than having to extract some elements of data to address them as rows and columns.” Many organizations instead enrich their data through the ETL process: extract, transform and load. But that’s an extra layer the data must go through, and it doesn’t necessarily train the data as it’s ingested. With NoSQL and multimodel, you get greater agility and speed in your data by avoiding that extra step, as well as delivering with scalability, Washburn said. That’s far more effective in achieving data readiness for AI or machine learning, he added. It also enhances transparency, which aligns the AI goals for many federal agencies, not least of which is the DoD’s “Ethical Principles for Artificial Intelligence.” Those principles require AI, as well as the data it uses, to be responsible, equitable, traceable, reliable and governable. The goal of the principles is to avoid bias in the data, build trust in the AI models and their decisions, and essentially avoid a “black box AI” situation in which the decision-making process cannot be reverse-engineered or understood by its operators. “No system that brings data in should ever be allowed not to return the data in the way that it was given. And I think that’s a perspective thing that the government needs to get ahold of,” Washburn said. “Because if there is a system that’s consuming data, a pure audit alone is required. If a system is consuming data and the origination is lost forever as it makes its way through a system, process or application, then how do I know what occurred to my data if I don’t know what I originally had?” That’s why building trust through provenance and lineage is so important, he said. Some systems change, add to or curate information. But any data that goes into a system should be able to be easily extracted in the same form that it was ingested. That’s the first layer of trust and adherence to those five DoD principles, Washburn added. The second layer is being able to track and trace what goes into the system. AI/ML systems may have to adjust data for context. For example, when ingesting names from documents, a system might need to understand that some cultures place family names before given names. In those cases, it might not be appropriate to address a person by their first name. An equitable system should be able to identify these cases and adjust appropriately, and a transparent system should be able to explain when and why it did so, Washburn said. “We may have that understanding as a human, based on the data that we know has been brought in. But the machine’s not going 8 PROVIDED BY MARKLOGIC Ensuring data transparency into AI processes Monitoring AI assets No system that brings data in should ever be allowed not to return the data in the way that it was given. — MarkLogic Federal’s Bill Washburn

Enable Your Data Strategy for Mission Success SIMPLIFY COMPLEX DATA & ACHIEVE DATA AGILITY LEARN HOW WE SUPPORT THE DOD & IC AT: MarkLogic.com/national-security Get a faster, trusted way to unlock value from complex data – and achieve data agility to meet mission requirements now and in the future. Our unified data platform combines a multi-model database and semantic AI technology to provide a comprehensive data layer with strong security, integration, and scale. PROVIDED BY MARKLOGIC 9 to have that understanding until you tell it. And I think the approach that you want to take to make that fast and consistent is to tell the machine the same way it’s applied to a machine learning model. And applying metadata is a quick way to do just that.”

Through an artificial intelligence set of principles, the Biden administration is urging agencies to move on from talking about AI and instead start using it and other automated tools more widely in day-to-day work. The “Blueprint for an AI Bill of Rights” outlines what agencies should do to ensure AI tools designed, developed and deployed — in and out of government — align with privacy rights and civil liberties. The administration, as part of these efforts, is also working on new federal procurement policy and guidance, to ensure agencies buy and implement AI and automation tools that are transparent and free of bias. Sorelle Friedler, assistant director for data and democracy at the White House Office of Science and Technology Policy, said that the blueprint is “putting the weight of the White House” behind a policy area that’s provoked a lot of conversation but hasn’t led to widespread implementation across government. “We are not really breaking new ground but adding to the conversation and helping us move the conversation forward, from principles into practice,” Friedler said at a Brookings Institution event about the bill of rights. The Biden administration also released a technical companion to the blueprint that serves as a roadmap for implementing transparent and accountable AI tools in government. “We are also trying to live up to that across the federal government,” Friedler said. The nonbinding policy document puts a governmentwide focus on automated systems White House aims to accelerate government AI use through ‘bill of rights’ BY JORY HECKMAN We are not really breaking new ground but adding to the conversation and helping us move the conversation forward, from principles into practice. — Sorelle Friedler, Assistant Director for Data and Democracy, OSTP White House releases AI roadmap 10 FEDERAL NEWS NETWORK EXPERT EDITION: DELIVERING ON ARTIFICIAL INTELLIGENCE’S POTENTIAL

_Hlk122531495

0/0

匹配案例每页限制结果 1 个

123456789101112131415161718192021222324252627282930313233/ 33100%实际尺寸适合宽度适合高度适合页面自动介绍杂志连续的

Delivering on artificial intelligence’s potential Insights from • Army • Brookings Institution • CISA • DoD • GSA • National Geospatial-Intelligence Agency • National Reconnaissance Office • National Technical Information Service • Office of Science and Technology Policy • Upturn EXPERT EDITION BROUGHT TO YOU BY

TABLE OF CONTENTS Investing in federal AI skills …………………………. 4 Achieving data readiness for AI/ML …….. 7 Accelerating federal AI adoption …………………… 10 Driving AI to the tactical edge …………………. 14 Securing federal supply chains ………………. 17 Implementing AI identity proofing …………. 20 Streamlining Army selection board processes ……………………… 23 Embracing the art of the possible ………………….. 25 Expanding IC partnerships with industry ……………………….. 28 Moving up the AI maturity curve …………….. 31 Can AI let your teams use their brains better? That headline is not meant to be demeaning. It’s a serious question, and one for which the answer appears to be a resounding “yes” — from both federal technology leaders and industry developers of artificial intelligence and machine learning technologies. Most people agree that one of AI/ML’s greatest potential benefits in government is that it can free data scientists and subject matter experts from the tyranny of clerical and mundane tasks. They then can use their expertise to wrestle with mission challenges and other complex demands. There are more than a few examples. Here are two: • Consider federal supply chain risk management efforts. “There really is a very shallow pool of subject matter experts out there in this area. Because that pool is so shallow, we have to turn to automation to help us,” shares Brian Paap, cyber supply chain risk management lead at the Cybersecurity and Infrastructure Security Agency. • Think about decision-making on the battlefield. “At the edge, the ability for things to be deployed and automated versus needing large teams of people to come in and to do those deployments is going to be critical. Warfighters are very talented folks, but they may lack the IT talent at the edge to do this. And that’s where we think automation can be helpful,” suggests Jim Keenan, vice president for DoD at Red Hat. In this ebook, we share strategies and tactics for accelerating and maturing federal AI/ML initiatives, along with details about implementing new technology tools, establishing appropriate guardrails and developing metrics for success. In the 10 articles, you will discover advice and insights from multiple agencies as well as AI leaders in industry. We hope it will help your organization mature its own use of AI/ML as you strive to make smart decisions faster by relying on data. Vanessa Roberts Editor, Custom Content Federal News Network FEDERAL NEWS NETWORK EXPERT EDITION: DELIVERING ON ARTIFICIAL INTELLIGENCE’S POTENTIAL 3

With the National Artificial Intelligence Initiative Act hitting its two-year anniversary, federal leaders are looking at more ways to invest in their workforces to better implement AI tools. Although understanding the talent and skills needed to better take advantage of AI is key, there are additional barriers to implementation. “What we have is a large number of federal agencies that are struggling with antiquated architectures and a lack of skills and talent,” said Chakib Chraibi, chief data scientist at the National Technical Information Service at an ATARC event on implementing AI. AI can play a crucial role for federal agencies, if they are able to implement it effectively. That means creating responsible guardrails like privacy, transparency and fairness in the use of AI, Chraibi said. At NTIS, “it’s helped us make evidence-based decisions, improve on customer experience, perform intelligent automation, enhance data privacy and ethical data practices, as well as strengthen our cybersecurity systems,” he said. Some agencies are already aiming to make more internal investments to boost their workforces’ understanding of AI, as well as train current employees on best practices. The Army, for instance, is looking to do more internal upskilling and recruiting, while also working to maintain industry partnerships. “It’s key for anybody embarking on an AI journey to know and understand your organization’s mission and how AI can enable it,” said Army Forces Command Chief Data Officer Jock Investing in workforce AI capabilities Agencies should make internal workforce investments to improve AI implementation, experts say BY DREW FRIEDMAN 4 FEDERAL NEWS NETWORK EXPERT EDITION: DELIVERING ON ARTIFICIAL INTELLIGENCE’S POTENTIAL

Padgett at the ATARC event. “You don’t always want to outsource your data and AI talent, so invest in your people upfront.” The Army also wants to add to the internal team that deals directly with AI-related work, Padgett said. Although the Army can make direct hires for software developers and data engineers, the service still relies heavily on the private sector to hire data scientists. “Data scientist talent is very weighted on the industry side right now. What I do see happening over the course of several years is that scale will end up balancing itself out to some degree, as DoD as a whole starts taking on the training tasks, new skill sets [and] upskilling,” Padgett said. For the Defense Department overall, Jaret Riddick, DoD’s acting principal director for trusted AI and autonomy, said that diversity, equity, inclusion and accessibility also play a role in the recruitment process. The Navy, for example, recently invested roughly $27 million to expand its Historically Black Colleges and Universities/Minority Institutions Program, Riddick said. These types of investments help DoD with “expanding the aperture to look for talent.” “Down the road, there will be a critical need to grow the talent base and to maintain an eye on the capacity of the industrial base in the future, to produce these technologies that we’ll need,” Riddick said at the ATARC event. AI is not the only area where DoD is looking to expand its connections with HBCUs. In June 2022, DoD and the Air Force created and funded a new research institute, partnering with 11 minority institutions to create the organization. Along with these types of minority institution partnerships, DoD is adding other industry partnerships as well. “We are promoting the growth of new companies, startups and small businesses [and] we are, of course, engaging with the traditional players,” Riddick said. To best implement and use AI, at least some understanding of the technology is necessary at all levels of an agency’s workforce, Chraibi said. By assessing the internal resources and skills Data scientist talent is very weighted on the industry side right now. What I do see happening over the course of several years is that scale will end up balancing itself out to some degree, as DoD as a whole starts taking on the training tasks. — Jock Padgett, Chief Data Officer, Army Forces Command DoD expands investments in HBCU programs FEDERAL NEWS NETWORK EXPERT EDITION: DELIVERING ON ARTIFICIAL INTELLIGENCE’S POTENTIAL 5

that are currently available, agency leaders can then focus on upskilling and training where it’s needed. They can also identify what support they may still need to obtain from external sources. “It’s important to have leadership understand this technology, understand what are the needs, what are the requirements, and of course, supply the skills and resources that are needed to be successful,” Chraibi said. It’s important to have leadership understand this technology, understand what are the needs, what are the requirements, and of course, supply the skills and resources that are needed to be successful. — Chakib Chraibi, Chief Data Scientist, NTIS 6 FEDERAL NEWS NETWORK EXPERT EDITION: DELIVERING ON ARTIFICIAL INTELLIGENCE’S POTENTIAL

How agencies can achieve data readiness for AI Agencies across the federal government are embracing artificial intelligence and machine learning, but the first challenge they often run into is preparing their data. Data readiness involves having a managed concept of data that allows the data to understand what model of AI is being applied: machine-machine, machine-human or human-machine. Having data readied will lead to greater speed, agility and transparency in the data. Enrichment of metadata is a good way to accomplish this. It lets the machine understand more about the data, whether it receives it from a human or another machine, and it allows humans to know more about data output by a machine. That can answer a lot of analytical needs for additional information about the data, as well as supporting the query layer of AI systems. One effective way to do this is to store the metadata alongside the data itself. “Not only do you get traceability with the data being alongside your metadata, but you also get that information in a very fast and effective way,” said Bill Washburn, chief program officer at MarkLogic Federal. “If you’re doing a geospatial search and you’re looking for a plot on a map, or if you’re looking for a section of information through a visual acuity — like not just a map but maybe a video, that information being stored alongside means that I only have to search for that information,” he added. “There is a great advantage in having those two things be symbiotic.” That’s especially helpful when dealing with unstructured data, which the government deals with in droves. The Defense Department and law enforcement agencies are applying AI to video analytics. The intelligence community, Interior Department and agencies like the National Oceanographic and Atmospheric Administration make frequent use of maps and satellite images. And most agencies, including the Veterans Affairs Department and the National Archives and Records Administration, are working hard to digitize paper records. “An advantage of NoSQL is you don’t have to convert images to text. You don’t have to change Managing unstructured data An advantage of NoSQL is you don’t have to convert images to text. You don’t have to change it. You don’t have to wait for it to be modeled. — Bill Washburn, Chief Program Officer, MarkLogic Federal PROVIDED BY MARKLOGIC 7

it. You don’t have to wait for it to be modeled,” Washburn said. “That can be modeled as your data is, rather than having to extract some elements of data to address them as rows and columns.” Many organizations instead enrich their data through the ETL process: extract, transform and load. But that’s an extra layer the data must go through, and it doesn’t necessarily train the data as it’s ingested. With NoSQL and multimodel, you get greater agility and speed in your data by avoiding that extra step, as well as delivering with scalability, Washburn said. That’s far more effective in achieving data readiness for AI or machine learning, he added. It also enhances transparency, which aligns the AI goals for many federal agencies, not least of which is the DoD’s “Ethical Principles for Artificial Intelligence.” Those principles require AI, as well as the data it uses, to be responsible, equitable, traceable, reliable and governable. The goal of the principles is to avoid bias in the data, build trust in the AI models and their decisions, and essentially avoid a “black box AI” situation in which the decision-making process cannot be reverse-engineered or understood by its operators. “No system that brings data in should ever be allowed not to return the data in the way that it was given. And I think that’s a perspective thing that the government needs to get ahold of,” Washburn said. “Because if there is a system that’s consuming data, a pure audit alone is required. If a system is consuming data and the origination is lost forever as it makes its way through a system, process or application, then how do I know what occurred to my data if I don’t know what I originally had?” That’s why building trust through provenance and lineage is so important, he said. Some systems change, add to or curate information. But any data that goes into a system should be able to be easily extracted in the same form that it was ingested. That’s the first layer of trust and adherence to those five DoD principles, Washburn added. The second layer is being able to track and trace what goes into the system. AI/ML systems may have to adjust data for context. For example, when ingesting names from documents, a system might need to understand that some cultures place family names before given names. In those cases, it might not be appropriate to address a person by their first name. An equitable system should be able to identify these cases and adjust appropriately, and a transparent system should be able to explain when and why it did so, Washburn said. “We may have that understanding as a human, based on the data that we know has been brought in. But the machine’s not going 8 PROVIDED BY MARKLOGIC Ensuring data transparency into AI processes Monitoring AI assets No system that brings data in should ever be allowed not to return the data in the way that it was given. — MarkLogic Federal’s Bill Washburn

Enable Your Data Strategy for Mission Success SIMPLIFY COMPLEX DATA & ACHIEVE DATA AGILITY LEARN HOW WE SUPPORT THE DOD & IC AT: MarkLogic.com/national-security Get a faster, trusted way to unlock value from complex data – and achieve data agility to meet mission requirements now and in the future. Our unified data platform combines a multi-model database and semantic AI technology to provide a comprehensive data layer with strong security, integration, and scale. PROVIDED BY MARKLOGIC 9 to have that understanding until you tell it. And I think the approach that you want to take to make that fast and consistent is to tell the machine the same way it’s applied to a machine learning model. And applying metadata is a quick way to do just that.”

Through an artificial intelligence set of principles, the Biden administration is urging agencies to move on from talking about AI and instead start using it and other automated tools more widely in day-to-day work. The “Blueprint for an AI Bill of Rights” outlines what agencies should do to ensure AI tools designed, developed and deployed — in and out of government — align with privacy rights and civil liberties. The administration, as part of these efforts, is also working on new federal procurement policy and guidance, to ensure agencies buy and implement AI and automation tools that are transparent and free of bias. Sorelle Friedler, assistant director for data and democracy at the White House Office of Science and Technology Policy, said that the blueprint is “putting the weight of the White House” behind a policy area that’s provoked a lot of conversation but hasn’t led to widespread implementation across government. “We are not really breaking new ground but adding to the conversation and helping us move the conversation forward, from principles into practice,” Friedler said at a Brookings Institution event about the bill of rights. The Biden administration also released a technical companion to the blueprint that serves as a roadmap for implementing transparent and accountable AI tools in government. “We are also trying to live up to that across the federal government,” Friedler said. The nonbinding policy document puts a governmentwide focus on automated systems White House aims to accelerate government AI use through ‘bill of rights’ BY JORY HECKMAN We are not really breaking new ground but adding to the conversation and helping us move the conversation forward, from principles into practice. — Sorelle Friedler, Assistant Director for Data and Democracy, OSTP White House releases AI roadmap 10 FEDERAL NEWS NETWORK EXPERT EDITION: DELIVERING ON ARTIFICIAL INTELLIGENCE’S POTENTIAL

Also read: