[2:48] Tim’s Career Roles In and Outside of Microsoft
[6:35] Making Correct Predictions In the Early Days of the Internet
[11:55] Tim’s Predictions for the Future of AI
[17:35] Moore’s Law, AI, and the Exponential Computing Power of Microprocessors
[22:02] Stumbling Blocks to Fueling Processing Power
[28:00] The Ethical Questions of AI
[18:12] “If Moore’s Law continues to execute for three more years… our computers can compute at the speed of the human brain.”
[32:24] “You end up having to combat AI with more AI.”
[37:13] “With all of the compute power that we know we are getting, we’ll fix the real problems.”
For better and for worse, the future of AI is coming at us faster than ever. In this episode of the Absolute AI podcast, Melody welcomes Tim Huckaby. Tim is an industry luminary focused on AI, Computer Vision, Machine Learning, AR/MR, Data Visualization, & Edge Computing. Tim has over 35 years of technology experience including 25 years with Microsoft where he worked on some famous and not so famous Microsoft Server products in the late nineties and learned how to build scalable enterprise software. He founded InterKnowlogy in 1999 and Actus Interactive Software in 2011 which became VSBLTY in 2015, and now is an AI consultant for Innodata. Tim is an accomplished keynote speaker and writer including authoring 3 books and hundreds of articles on topics ranging from Emerging Experiences in Artificial Intelligence, Computer Vision, Voice Recognition, and futuristic interfaces. Together they discuss Tim’s computing predictions that have already come to fruition, the ones he has for the future of AI, and the stumbling blocks of AI that will have to be overcome before it can truly become a problem-solving part of our everyday lives.
INSIGHTS FROM THE PIONEER OF THE SMART CLIENT REVOLUTION
A true pioneer of his field and never one to shy away from a technical discussion, Tim was once deemed the pioneer of the smart client revolution when he made a prediction about locally installed native applications dominance, including the use of apps on smartphones. What was he able to see regarding web applications that others were missing? Although he initially predicted that the internet would die under its own weight and band restrictions, he also recognized that true technology advances would have to happen on phones. 15 years later, he would argue that the prediction was both entirely correct and completely wrong.
PREDICTIONS FOR THE FUTURE OF AI
As Moore’s Law continues to execute and the exponential growth of microprocessors continues to double every two years, we move closer to the ability of CPUs to compute at the speed of the human brain. AI is transforming from its primitive abilities of common object recognition to identifying types of cancer, Covid, commercial products, and advanced recognition. And while this doesn’t mean that the robots are about to take over, it does mean that menial tasks will continue to be completed, much more effectively, by AI devices.
OVERCOMING STUMBLING BLOCKS TO DOING MORE
Data is the number one hurdle that has to be overcome when harnessing the real power of machine learning. Tim walks listeners through the greatest ideas that have never come to fruition because of access to data. Privacy laws, manual labor, and the sheer number of data points required to create some apps have severely limited their actualization.
ADDRESSING THE ETHICAL POINTS OF AI
Deep fake videos are a perfect example of the downside of AI capabilities. There will continue to be innovations that are never released to the general public because they are too dangerous, but ultimately, the negative capabilities of AI will be fought with even more AI. Similarly, quantum computing will be combated with more quantum computing. Advancements are only a few years away from successfully creating a quantum computer, but what does that mean for the future of AI?