As companies make the shift into cloud computing, there is always room for improvement, from computational waste to monopolized cloud computing. Leon Kuperman, Co-Founder, and CTO at CAST AI, told The Yield that computer waste was “a problem that I felt instinctively in our last startup and I couldn’t solve it, and then I saw other customers suffering with that same problem and I decided that we were going to start a company…to solve that problem for the world.”
[12:31] What are the possibilities for the future of machine learning?
[30:41] Government steps to shoring up infrastructure.
[38:05] The impact of machine learning on everyday consumers.
Technology is influencing and changing every aspect of our lives, and even more so with AI and machine learning, according to Kuperman. But something that is not often discussed is the waste generated by digital technology. This is something Kuperman has been fighting to eliminate with his company CAST AI. “Nobody cares about waste when everyone’s growing hand over fist, it’s not a problem to solve for today. We started solving this problem a couple of years ago because now we’re going to see this inflection point and this shift”
This is where AI and machine learning play a role. AI and machine learning can help get rid of all of the ‘gunk’ in between and let engineers focus on creativity, Kuperman said. Machine learning utilizes predictive analysis to predict traffic patterns, anticipate peaks and valleys, the complex cyclical patterns in the cloud, and ultimately determine, without any human intervention, the types of computers and infrastructure customers need to effectively get business done.
Sign Up On Yieldstreet
The problem with consumer created cloud waste is, “partially kind of an architecture problem and it’s partially an accountability problem” according to Kuperman due to things like sandbagging. He said, “if you go through two or three layers of sandbagging all of a sudden, you have an application that needs 50 CPUs to run and it’s running on 2000 CPUs and that’s where all the waste is.” Kuperman went on to say the goal is to help people stop working on these low level infrastructure issues and bring their skill sets into the most creative path, which is delivering value to their business.
Another aspect of AI that Kuperman discussed was the touchpoints consumers hit in their everyday use of digital technology. The average consumer is already interacting with AI at hundreds of touchpoints throughout the day, seemingly with little concern over the privacy and data that is required to do so. Kuperman said, “I think they are interacting with AI far more than the average consumer knows. It’s everywhere from the movie that you’re being recommended to how your phone is suggesting that you leave now to make your dentist appointment or the typical workout that you do when you didn’t even put that in your calendar. Consumers have given up on the fact that there’s any privacy,” Kuperman pointed out. “I think very few people have an intrinsic expectation of privacy versus the convenience that the data that they provide gives them,” he added.
As for the next phase of AI and machine learning, Kuperman said “where you’re starting to see the next evolution of machine learning is when there are less rules in place…and start applying complete creativity”.
Take all of our content on the go with you and listen on demand no matter where you are. Subscribe on Apple Podcasts, Spotify, YouTube, or wherever podcasts are available so you don’t miss an episode.
Yieldstreet provides access to alternative investments previously reserved only for institutions and the ultra-wealthy. Our mission is to help millions of people generate $3 billion of income outside the traditional public markets by 2025. We are committed to making financial products more inclusive by creating a modern investment portfolio.