“Gartner's Key Initiative Leader for AI discusses the importance of responsible AI development.”
Gartner's Key Initiative Leader for AI, Anthony Mullen, explains the importance of responsible AI in a world where more organizations are implementing AI systems.
In this video, Anthony Mullen, the Key Initiative Leader for AI at Gartner, discusses the topic of responsible AI. With the barriers to using AI becoming lower, more and more organizations are interested in implementing AI systems. However, this also means that the risk plan for AI has massively increased, making it essential to take an approach that is explainable, understandable, reduces bias, and ensures privacy when developing systems that interact with people, economies, and society. Mullen explains that there are technical approaches that we can take to ensure responsible AI, but it's also essential to take a human-centric approach since these systems are developed with humans to meet human goals. However, human weaknesses and biases can also be put into these systems. Responsible AI is not just a civic duty but also enables us to develop more effective systems, which can influence how society operates, from politics to social media and human values.
Mullen goes on to discuss the various risks associated with AI, from small mistakes in image classification to influencing how society operates. The risks are complex to understand, and controlling them can be challenging, especially with technologies like neural networks, which are essentially a black box. Most of the bias comes from the data that we train these models on. For example, if computer vision models are only trained using white middle-aged men, the systems will be biased towards that. To mitigate these risks, education of those developing and consuming AI systems is crucial, along with exploring risks and developing new language to detect bias and protect privacy.
Mullen also explains that there are already well-developed ethical frameworks around the world, and instead of reinventing the wheel, we should take these frameworks and see how they would work within the context of AI. We need to develop the language of risk for AI and work with governments, international bodies, staff, customers, and partners to set up safe open exploratory ecosystems.
In conclusion, Mullen emphasizes that the best practices for responsible AI are to involve legal and security teams at the very beginning of projects, set up a workflow from a team outside of the development team, and involve an Ethics board in the testing of use cases. These practices not only adhere to responsible AI principles but also enable the development of more effective systems. Watch this video to learn more about responsible AI and how it can benefit society.
#cxotvnews #technology #business #digitaltransformation #gartner #artificialintelligence #innovation #security #securitysolutions