I was recently featured as a guest speaker at the DNX® Venture Webinar – ZERO-TIME SUPPORT A UTOPIA OR WITHIN REACH. The Moderator Eva Nahari asked me the following question that I thought of sharing with you:
Question: Is AI actually economically feasible for support?
My answer: This is a complex question. In my opinion, our industry is making significant improvements in the following three areas that are making AI more feasible economically:
– Onboarding Cost: We have made significant progress from the IBM Watson days when it would take months/years and an army of professionals to train a model. Today with our solution we are onboarding a customer in about a month with our pre-trained model.
– Maturing AI Models: We are moving from Traditional Models to Deep Learning Models that are more complex but capable of solving even more complex issues.
– Processing Cost: we are moving from using expensive TPUs to relatively cheaper GPUs. We are using Nvidia GPUs for our processing. I am hearing that some companies have started using CPUs for their processing.
If interested, you can watch the entire webinar on our website: https://lnkd.in/ecDDR9EZ
As always, I would love to hear your thoughts.