Jeffrey Funk Dissected the Problem with AI Breakthroughs and the Power of Augmentation
Have you wondered why some hyped AI breakthroughs have yet to come true? What are the reasons they have failed? How can we capitalize on the huge potential and capabilities of artificial intelligence for humans and machines to work well together?
Dr. Jeffrey Funk, a former professor at the National University of Singapore and now a technology consultant, attended Worldwide AI Webinar and answered these questions in his keynote about AI breakthroughs and why we need more augmentation.
Two problems with AI breakthroughs
According to Dr. Funk, predicted AI breakthroughs such as autonomous vehicles or predicting crimes using AI need 99.9999% reliability in safety-critical functions.
This could take decades because it took Google, an already mature website, 25 years to reach 99.999% uptime. Research from the University of Chicago reported that algorithms can only predict future crimes with 90% accuracy when it merely explains historical data. Additionally, two Chinese researchers claimed that computer algorithms could only predict criminals using photos with 89.5% accuracy.
Facial recognition systems used by police to identify criminals are also not very good since according to the research, in some cases, they would just set the confidence level to 50% and see what happens.
While Dr. Funk shared lots more examples, the bottom line is that machine-powered and machine-dependent tools are not yet developed enough to have complete autonomy and replace human workers.
There are two reasons why:
A lot of systems don't work as well in the field as they do in the lab where they were trained
It takes many computing resources to increase accuracy and reduce the error rate
Augmentation is the way to go
Dr. Funk claimed that augmentation could be working in many industries including radiology. He cited a 2022 research paper concluding that AI and doctors could work better together to detect breast cancer.
The research also reported that AI could do small tasks for doctors so they could save lots of time and focus on important ones. For this to happen, augmentation is much needed.
Jeffrey Funk ended his keynote by stating that AI would diffuse slowly and emphasizing the possibility and power of augmentation in the future.