Industrial-Grade Machine Learning at GE— An Interview with Joshua Bloom
In this episode of the Masters of Data podcast, I speak with a leader in the world of machine learning. Joshua Bloom is Vice President of Data and Analytics at GE Digital, where he serves as the technology and research lead, bringing machine learning applications to market within the GE ecosystem. Previously, Joshua was co-founder and CTO of Wise.io, which was acquired by GE Digital in 2016. Since 2005, he has also been an astronomy professor at the University of California, Berkeley, where he teaches astrophysics and Python for data science. Needless to say, Joshua is someone who is more than capable of speaking about the topic at hand. Joshua sits down me to discuss the events that led up to Wise.io’s genesis and its place in the world of technology but also more broadly about how machine learning is developing, specifically as it relates to the industrial world.
Joshua starts by offering a little insight into his background and how his training in astronomy set the stage for coming into the world of machine learning and ultimately to build Wise.io. As Joshua explains, after graduation, “I started getting interested, not just in the objects that we were studying, but in the ways in which we were studying them…I started thinking about what was going to happen in the astronomy world when we had 10x, 100x, 1,000x more data, I got pretty scared at some level. But, I also got pretty excited because it felt like there was some intellectual white space that me and my colleagues were not yet grabbing onto.” And what role did his background in astronomy play in his growing interest for learning and problem-solving? “I think like all good astronomers throughout history, they look around for other tools that other people are using…Astronomers are pretty good at pooling tools and toolkits and approaches from other places.” And this was the initial emphasis of Wise.io; taking the tools of the other industries around them and applying them in a unique way to solve a different problem.
So how did Wise.io start? Well as Joshua explains, “I started…thinking about how machine learning could be useful in my own world, and in particular learning how to scale out of some of the problems that I saw coming down the pike. And, that was really the genesis of how I got into data science more broadly, how we wound up starting [Wise.io].” He continues, “What we wound up coming to realize is that to succeed in this world, to bring machine learning in some sense to the masses, somehow we needed to pick a vertical. We needed to pick a specific set of use cases with a targeted set of users and a very clear buyer in mind that would benefit from the AI that we had inside.” “So, what we wound up doing in this company…was build a set of applications on top of Salesforce and Zendesk to help with customer support. And, what we wound up seeing there is a little bit like what we’re seeing on the astronomy side, where you had deep domain experts who knew, in this case, a product line incredibly well, and were the front lines of a company succeeding or failing based on their interactions with customers,” he notes. As Joshua explains, that ideology was the start, but has been translated in different ways across industry. But aside from Wise.io’s start Joshua and I also take the time to discuss pressing ideas like bias and machine learning.
While bias is a hot topic in the machine and data world, there has to be clear direction on how it’s understood. So how do we tackle this idea of bias? Well, as Josua notes, “Bias obviously is extremely dangerous in the consumer world and the person facing world. Essentially, when you model data from the past, you’re building in all of the decisions that were made either consciously or unconsciously…But, it has to get looked at algorithmically. And, I think one of the exciting frontiers of machine learning these days is understanding at the deep theoretical level, how we can understand bias, quantify bias, and ultimately protect against that.” So while it may be somewhat inevitable, Joshua believes that using machine learning can actually help to minimize the impact bias has. But what is the role of machine learning in industry? “When you think about data science, and you think about the machine learning in the world that we’re in, in the industrial internet of things…the data that’s coming off of those machines, while they aren’t perfect exemplars of that physical object, they are pretty good proxies for what’s happening with that physical object.”
Aside from the theory of how machine learning functions in the world of industry, Joshua also offers specific context about how it is proving to be valuable to consumers and industries. “When GE sells a jet engine to a company, we’re also selling some level of assurance about the quality of that object going forward over decades time. This is not an iPhone where in two years from now you throw the thing out and you buy a new one, right? These have to live for a substantial amount of time and they always need maintenance,” he explains. “Our job as we’ve wound up building an application internally within GE and with their aviation partners, is to build a system that allows the experts to focus on the hard problems. So, something that obviously needs to be looked at, go for it, right? We’ll just help them automate that, or we’ll help them at least make a decision more quickly…And, if it’s done in a nonintrusive way, it becomes accepted and becomes sort of a partner in the decision making process.” So by using machine learning to help aid with the basic (yet often overwhelming) task of maintenance, industries and companies can help better focus their time and attention as it’s best needed for the global good.
Outbound Links & Resources Mentioned
Masters of Data Episode — https://www.sumologic.com/podcast/industrial-grade-machine-learning-joshua-bloom/
Learn more about Joshua:
https://www.wise.io/meet-the-team/
Connect with Joshua on LinkedIn:
Follow Joshua on Twitter @profjsb
Learn more about Wise.io:
Follow Wise.io on LinkedIn:
https://www.linkedin.com/company/wise-io/
Takeaways
- All good astronomers throughout history look around for other tools that other people are using. The famous example of course is Galileo instead of taking a telescope and pointing at the horizon because it had been invented for military purposes, to look for ships coming over the horizon.
- Astronomers are pretty good at pooling tools and toolkits and approaches from other places.
- Major discoveries happen when people figure out ways to cross over domains.
- Having a physics background, or any sort of more broad training and physical sciences, helps you think about the ways in which you tackle problems from a first principle’s perspective.
- Physical sciences enable you to ask questions like, “What are the irreducible components of this problem that I’m looking at? And, how do I attack them individually? Can they be attacked individually, or is it such a complex issue that you have to attack them holistically?”
- When you think about data science, and you think about the machine learning in the world that we’re in, the data that’s coming off of those machines, while they aren’t perfect exemplars of that physical object, they are pretty good proxies for what’s happening with that physical object.
- To succeed in this world, to bring machine learning in some sense to the masses, somehow you need to pick a vertical. You need to pick a specific set of use cases with a targeted set of users and a very clear buyer in mind that would benefit from the AI.
- Wise.io built a set of applications on top of Salesforce and Zendesk to help with customer support.
- They aimed to solve the problem where you have deep domain experts who know a product line incredibly well, and are the front lines of a company succeeding or failing based on their interactions with customers.
- Bias is extremely dangerous in the consumer world and the person facing world. Essentially, when you model data from the past, you’re building in all of the decisions that were made either consciously or unconsciously.
- Bias it has to get looked at algorithmically. And, I think one of the exciting frontiers of machine learning these days is understanding at the deep theoretical level, how we can understand bias, quantify bias, and ultimately protect against that
- When GE sells a jet engine to a company, it’s also selling some level of assurance about the quality of that object going forward over decades time. This is not an iPhone where in two years from now you throw the thing out and you buy a new one. These have to live for a substantial amount of time and they always need maintenance.
- What GE has is a number of monitoring services that look at the health and quality of something like a jet engine. And, they give feedback to the parent companies saying, here are our insights about the engines that are flying. And, ultimately it’s up to them to sort of make decisions about what they want to do for maintenance perspective.
- GE’s job is to build a system that allows the experts to focus on the hard problems. They just help automate that, or help them at least make a decision more quickly.
- If it’s done in a nonintrusive way, it becomes accepted and becomes sort of a partner in the decision making process.
- The vision for the future is using machine learning in the private sector to see how can we build machine learning models where both sides, or multiple parties can wind up benefiting, but where there is no data leakage.
