Keynote speakers highlight change

Garrett Neese/Daily Mining Gazette Phil Bourne delivers a keynote address during the second day of Michigan Technological University’s computing showcase Tuesday.

HOUGHTON — Change was on the minds of two speakers Tuesday at the erstwhile Michigan Mining School.

Dianne Marsh, director of content security for Netflix, and Phil Bourne, dean of the University of Virginia’s school of data science, delivered keynote addresses Tuesday during the second day of Michigan Technological University’s computing showcase.

There are two types of change, said Marsh, a Tech graduate: change that happens to you, and change you make happen.

“It’s always uncomfortable,” she said. “With that discomfort, though, is where we get growth.”

Netflix founder Reed Hastings had always envisioned the company as a streaming service, but knew it would be unfeasible with late-90s internet speeds, Marsh said. As a waystation, he founded the company as a DVD-by-mail service, where it built the market that later enabled it to take on streaming.

Garrett Neese/Daily Mining Gazette Dianne Marsh delivers a keynote address during the second day of Michigan Technological University’s computing showcase Tuesday.

Fast-moving DVDs put pressure on the postal service. The sorting machines at the post office put another kind of pressure on the DVDs, leaving many of them cracked or scratched.

So Netflix figured out improvements to the sorting machine, which it then passed back to the postal service.

“You have to sometimes do things to affect the ecosystem in order to be able to do the things you want to do,” she said.

Another part of Netflix’s success has been knowing what steps not to take, Marsh said. The company had developed a streaming device. On the verge of going to market, Netflix pulled it back, worried a Netflix-branded device would make partners worried about not being on an equal footing. Instead, it was spun off into Roku.

“It would probably have been a good business model to have that device, but it was better not to have it,” she said.

At the end of Marsh’s talk, each table had a brainstorming discussion about both kinds of changes, and what their impacts were.

The changes going on in academia with digital transformation are bigger than anything Bourne has seen in a career of more than 40 years.

“We’re doing projects with religious studies, we’re doing projects with economics, we’re doing projects with biomedicine, we’re doing projects with fintech — all of them driven, obviously, by a lot of data,” he said.

Bourne pointed to the way the mapping of the human genome had shaped biomedicine, spawning new fields and a $600 billion industry. It wasn’t overnight: at one point, Bourne said, many people in computational biology who made the leap from academic to industry had circled back when early timetables proved too optimistic. By 2015, partnerships were growing between academia and experimenters. Sometime soon, he said, computation will be the primary driver of health care.

“In my mind, there’s no question that’s already started to happen,” he said.

UVA has embraced the “4 + 1” model for data science. A circle of four areas — value, systems, design and analytics — surrounds practice, which connects all of them at the center. The biggest emphasis is on value, where they explore the tension between what data science can do and its potential downsides, Bourne said.

The use of electronic health records brings privacy considerations with it, but also opens up new avenues for treating patients. As an example, he cited a trauma surgeon he knew who wanted his help with data science techniques to reduce the number of patients who died after traffic accidents. The surgeon used crash records from the Department of Motor Vehicles to develop a system that will correlate the type of damage seen in the crash to the type of injuries the people in the vehicle are likely to have. Emergency responders would send a photo to the ER, which would then be on the lookout for particular internal injuries.

“To me, this is the epitome of what data science should be doing,” he said.

Other sessions Tuesday included health information and infrastructure, machine learning and counterterrorism, and financial technology.


Today's breaking news and more in your inbox

I'm interested in (please check all that apply)
Are you a paying subscriber to the newspaper? *

Starting at $2.99/week.

Subscribe Today