In this interview, O’Reilly Foo Camp 2019 Dean Wampler, director of evangelism at Anyscale.ion, talks about transferring artificial intelligence and machine learning to real-time production environments.
Highlights of the interview include:
Facilitating the transition from research to production in a robust way brings with it a number of complications, including governance, GDPR, and traceability rules. Given the importance of traceability, he provides an example: “If I introduce a model that makes credit card authorizations and I constantly reject someone’s card, and they come to say,‘ I am a member of a minority group and you constantly reject my charges. Are you prejudiced against me? ‘or something like that, I need to know exactly what model was used and how it was trained. In a real production environment, all sorts of logistical issues need to be addressed. “(01:15)
In some cases, artificial intelligence and machine learning techniques are used to improve existing processes rather than solve new problems. Wampler used car loan approval as an example: “It took about a day to get a car loan, and it worked. You could just go back to the dealer the next day and dream of your beautiful car that night, but you don’t have it. Companies like Capital One have got it [loan approval process] seconds. You can access the app and get approval for the loan immediately. So, it’s not something was done in real time, but it changed the world, changed their business ability to do so. There are many such practical examples. “(02:22)
Wampler also discussed his personal interest in climate change and how individuals and companies can use artificial intelligence and machine learning tools to make a more significant impact than you might think. “I’ve learned that there are many small ways and big ways that come together when we work on things like that. One of the promises of tools like artificial intelligence is that it can automate human-level action in a way that wouldn’t be possible with real people. More specifically, Google organizations like this already use advanced analytics to reduce the amount of energy they use and use their machines more efficiently.Individually, things like that do not solve climate change, but they add up. small things that we can do and that add up, personal things like how we use energy, heat homes, prepare our meals and so on, always think carefully about how we do our work and how we can be effective by operating on these things, m thinking about how we can help our clients achieve this, and then thinking about ways we can make a more direct impact. “(04:20)