Author
The Pentagon was warned that it was missing the software revolution taking off in Silicon Valley. A bridge was built through the Defense Innovation Unit, but the turning point came when the Joint Artificial Intelligence Center (JAIC) confronted an uncomfortable truth: Machines were collecting more data than humans could process. We would quickly evolve the data science lifecycle and an AI revolution would ensue as the importance of AI in National Security would highlight its need to maximize capability relevance while also designing for ethical and secure uses of AI, addressing concerns about trust, agency, and the potential risks.
AI is and will continue to play a role in everyday decision-making and also disrupt the status quo by playing a pivotal role in everyday life and innovation. Most countries are pursuing AI advantages, some citizens with more pessimism than optimism. Comparing the outlook of U.S. citizens to citizens of China, the U.S. majority is more pessimistic over the future of AI while the Chinese majority holds much more optimism - as they push for a deeper merger of data and human thought such as embedded AI in robotics possibilities flume.
In a recent talk, Lt. Gen. John N.T. “Jack” Shanahan at Sandhills Community College highlighted the rapid advancement of AI, its lack of transparency, and the need for robust “design, testing, and validation”. With promise, these measures will reduce the learning curve of these complex systems, while building the trust of end-users is crucial when using AI to make critical decisions. How much of this is being today, is highly dependent of the investments in User Experience (UX) research and design to influence the requirements, with a human first mentality. You can’t unlock AI’s full potential without trust, but trust in full autonomous systems is held by only 60% of users. Perhaps there is a roadmap to full autonomy where users are infused in the design process to heighten transparency and adoption. Perhaps systems are made to respond to user preferences offering manual, semi-automatic or fully running agents that are capable of making decisions that impact the business, and meet users where they are with trust on a personal level. Consumers will not purchase or interact with systems they believe are unfair, unsafe or non-transparent. If an AI system is found to be biased, it can cause a significant erosion of brand loyalty and lead to legal exposure. Poor data quality and lack of clear data ownership are amplified by AI, creating decisions that are difficult to explain and impossible to defend.
Read more about 3 core pillars of AI-enabled data governance here.
