SNIM AI allows data from all connected machines to be pooled and used for retraining. The updated recognition system is then tested and verified by a human operator before being pushed back out to some of the machines in the field. The system is 'mission adaptive', meaning it only deploys relevant updates to specific machines. The need for rapid, responsive updates led Qylur’s CEO, Dr. Lisa Dolev, to develop SNIM AI, which she insists on keeping human oversight in the process to prevent AI from retraining itself in a way that exacerbates the problem.
Key takeaways:
- The US Air Force has awarded a phase I Small Business Innovation Research contract to Qylur Intelligent Systems to update AI systems in the field rapidly and efficiently using the Social Network of Intelligent Machines (SNIM AI).
- SNIM AI identifies problems within AI models and helps solve them with data collected by all the machines connected to the system, allowing for on-the-fly updates and retraining.
- AI model drift, where trained systems encounter difficulty with new situations, is a well-known issue in the business world, but in defense, these issues need to be rectified as soon as possible.
- While SNIM AI could be completely automated, Qylur CEO Dr. Lisa Dolev insists on human oversight in the process to ensure safety and reliability, especially as AI gains momentum in both business and military applications.