«Nous nous adaptons de façon encore plus réactive à l’évolution des menaces» : entretien avec Joël Barre

Sur le même sujet

• 

  29 mars 2019 | International, C4ISR, Autre défense

  DARPA: Using AI to Build Better Human-Machine Teams

  The inability of artificial intelligence (AI) to represent and model human partners is the single biggest challenge preventing effective human-machine teaming today. Current AI agents are able to respond to commands and follow through on instructions that are within their training, but are unable to understand intentions, expectations, emotions, and other aspects of social intelligence that are inherent to their human counterparts. This lack of understanding stymies efforts to create safe, efficient, and productive human-machine collaboration. “As humans, we are able to infer unobservable states, such as situational beliefs and goals, and use those to predict the subsequent actions, reactions, or needs of another individual,” said Dr. Joshua Elliott, a program manager in DARPA's Information Innovation Office (I2O). “Machines need to be able to do the same if we expect them to collaborate with us in a useful and effective way or serve as trusted members of a team.” Teaching machines social intelligence however is no small feat. Humans intuitively build mental models of the world around them that include approximations of the mental models of other humans – a skill called Theory of Mind (ToM). Humans use their ToM skill to infer the mental states of their teammates from observed actions and context, and are able to predict future actions based on those inferences. These models are built on each individual's existing sets of experiences, observations, and beliefs. Within a team setting, humans build shared mental models by aligning around key aspects of their environment, team, and strategies. ToM and shared mental models are key elements of human social intelligence that work together to enable effective human collaboration. DARPA's Artificial Social Intelligence for Successful Teams (ASIST) program seeks to develop foundational AI theory and systems that demonstrate the basic machine social skills necessary to facilitate effective machine-human collaboration. ASIST aims to create AI agents that demonstrate a Machine ToM, as well as the ability to participate effectively in a team by observing and understanding their environment and human partners, developing useful context-aware actions, and executing those actions at appropriate times. The agents developed under ASIST will need to operate across a number of scenarios, environments, and other variable circumstances, making the ability for them to evolve and adapt as needed critical. As such, ASIST will work to develop agents that can operate in increasingly complex environments, adapt to sudden change, and use observations to develop complex inferences and predictions. During the first phase of the program, ASIST plans to conduct experiments with single human-machine interactions to see how well the agents can infer human goals and situational awareness, using those insights to then predict their teammate's actions and provide useful recommended actions. As the program progresses, the complexity will increase with teams of up to 10 members interacting with the AI agents. During these experiments, ASIST will test the agents' ability to understand the cognitive model of the team – not just that of a single human – and use that understanding to develop appropriate situationally relevant actions. Full details on the program can be found in the Broad Agency Announcement (BAA) solicitation, which has been posted to the Federal Business Opportunities website, https://www.fbo.gov/index?s=opportunity&mode=form&id=9d4acf0aba98916288a541bd07810004&tab=core&_cview=1 https://www.darpa.mil/news-events/2019-03-21b

• 

  10 octobre 2024 | International, C4ISR, Sécurité

  OpenAI Blocks 20 Global Malicious Campaigns Using AI for Cybercrime and Disinformation

  OpenAI disrupts 20 malicious operations exploiting AI for cybercrime, including malware debugging, social media manipulation, and misinformation.

• 

  19 mai 2020 | International, Aérospatial, Naval, Terrestre, C4ISR, Sécurité

  Choosing the right commercial tech for government

  By: Meagan Metzger In today's crisis-stricken world, it is heartening to see leaders recognizing the importance of government support for innovative, private sector solutions to the problems facing the defense industry. For the Department of Defense, reforming policies and refocusing priorities so that commercial tech can be successfully implemented to support the defense industry's mission is essential. The DoD's endorsement not only encourages emerging tech startups to consider government compliance and scale in their business models from the very beginning; it also protects our national security — and service members in uniform — by putting the most innovative technology into play. But creating more opportunities for commercial tech companies to secure government contracts is only the beginning. For government agencies to successfully take advantage of innovative tech from the private sector, a few things need to happen — and the sooner, the better. First, the government needs to look beyond legacy contracts. As has been noted by venture capital leaders, the announced provisions of the coronavirus relief legislation, the CARES Act, “to streamline the Defense Department contracting process” currently apply only to contracts worth $100 million or more. This excludes emerging commercially successful tech companies that could have a significant impact at the government level. Separate, though related, are needed reforms to the Small Business Innovation Research program. The National Defense Authorization Act for fiscal 2020 provides additional SBIR flexibility for small businesses that are more than 50 percent owned by venture capital, but the DoD has yet to fully promulgate this new flexibility authority. Until eligibility standards are adjusted, the DoD is missing the chance to work with proven, VC-backed companies. Of course not all commercial tech companies are equipped to support government missions; and to ignore the importance of a rigorous evaluation process is even more harmful than ignoring commercial tech all together. Finding emerging tech is easy. Evaluating and equipping tech companies for success in government is hard, particularly when national security is a critical concern. The COVID-19 crisis has made it even more apparent that government agencies need to be able to implement tech solutions quickly and trust that they will perform as expected. A tech company with proven success in the private sector may draw the government's attention and show that it can deliver, but there are other equally important indicators to consider when determining if a company is capable of performing as expected at the government level. The Pentagon, like any government agency, must rely on data-backed advice and expertise to identify which commercial tech solutions are most likely to succeed in the federal market. Finding technology companies should not be a quantity play, but focus more on fit and quality. Moving fast requires working with private sector partners who have experience vetting tech companies for government contracts, which we've seen leaders do, like Space and Missile Systems Center's Air Force Col. Russell Teehan and the head of Air Force Program Executive Office Digital Steven Wert. Partners that are federally focused — with deep knowledge of government problem sets and missions — can identify which tech companies are viable technically and will be viable in the federal market. Assessing tech's viability requires specific experience evaluating a set of qualitative characteristics unique to this market, in addition to the typical “can they work with government" questions like: “Where is the code compiled?” Government agencies should also look to VCs and accelerators that can specifically guide tech companies through the government market contracting process and equip them to succeed in the long term. For instance, in 2019, the United States Air Force worked with Dcode to scout technology for the service's Multi-Domain Operations Challenge, and seven of the 30 finalists were companies that had completed the Dcode accelerator to prepare for success in the federal market. Supporting these tech companies requires more than just a singular contract award. To get over the “valley of death,” companies have to understand everything from compliance to how to staff, rework operational processes and market effectively, to name a few. There is no question that working with the right emerging tech companies is imperative for the DoD and other government agencies. But at a moment in history when time is particularly of the essence, there is no room for trial and error when it comes to identifying which tech companies can meet the government's specific needs. By working with private sector partners that have extensive government expertise and proven results, the DoD can confidently implement innovative technology that addresses its most critical needs in a time of crises and well into the future. Meagan Metzger is the founder and CEO of Dcode. She also serves on an advisory board for Booz Allen Hamilton, and another advisory board for the Defense Entrepreneurs Forum. She previously worked as chief operating officer of a mobile and cloud company, as well as chief strategy officer at an IT consultancy. https://www.defensenews.com/opinion/commentary/2020/05/14/choosing-the-right-commercial-tech-for-government/