As open source and open standards were -- naturally -- lauded at Boston's Red Hat Summit in May, Gahn Lane, CenturyLink's vice president of software alliances, echoed the message that open source "makes a lot of sense." He offered the caveat, however, that this sensibility depends upon a given organization's motivations and perceived-risk tolerance.
As CenturyLink Inc. (NYSE: CTL) was one of the main sponsors of Red Hat Summit this year, Lane was on site to sit down with Telco Transformation and offer his personal take on matters spanning big data, big video, and more. Here, in part one of this lightly edited Q&A, Lane talks about to what extent open source and open standards make sense for advanced analytics solutions like AI and IoT. Next up in part two, Lane discusses AI trends, network requirements for virtual reality (VR), and a key vertical that has heavy demands in both areas -- gaming.
Telco Transformation: From the DevOps perspective, one question that is constantly being asked about analytics and AI, especially in the startup community, is: "Do we write our own solutions, or do we use open APIs?" What's your take on that?
Gahn Lane: Well, since I'm at the Red Hat event [laughs], I'll certainly say that open source is a very popular path to take. If you are a startup, it makes a lot of sense because you're going to get more economies of scale and your solution's going to be more quickly accepted and recognized.
I don't know what software you're using on your laptop, but let's say you're using Windows, hypothetically. Windows is as common as bottled water, right? Whether you like it or not, whether you think it's the best or not is simply irrelevant. And as soon as we switch and you start using Google Chrome or something like that, now your PowerPoint doesn't work right or whatever, you know what I mean? It's the same thing with open source for AI, things like that. If you go with standards that are already in place -- because "open" is still "standard"; it's just an "interpretative dance" of standards -- you're going to more quickly market and sell your solutions.
So, for instance, at my company, if you came to me and you said, "Hey, I'm a startup and I've come up with this great AI solution for the Internet of Things…" If you tell me you based that on open standards versus something that you've invented yourself from the start from base layer, I'm probably going to be a little less comfortable with that, because that means I'm going to have to do a lot more investigation, testing, configuration, review, etc. If you say, "Hey, I built this on stuff I got at Red Hat…" then you automatically have credibility in the marketplace -- and our comfort level's going to be higher. A company my size doesn't want to spend too much time and money on somebody who has "a cool idea." It's too expensive. You don't have the time.
TT: What are you seeing here in terms of open-source needs for IoT verticals? Let's take smart cities, or example. Is there more of a need for open source and open standards because of the very nature of smart cities and municipal government? Does it matter much? And what are the advantages and disadvantages when we talk about the enablement of those solutions?
GL: We're more focused on -- I don't like calling them tier-2 or tier-3 cities -- but "non-NFL cities." So 50,000 people or below is where we typically have focused. And you [may ask] why. Well, because CenturyLink's the phone company in mostly rural areas, with the exception of a few places, like Seattle, Minneapolis, [Las] Vegas, Orlando. With the exception of those four, we're primarily in "non-NFL cities."
So the needs there, and the budgets there, are much tighter than they are in Manhattan, or one of the [other New York] boroughs, or Boston proper, or something like that. They're usually also tied to federal grant money. So, having said that, they're looking for, "How do I get there as quick and as inexpensive as possible because I only have x amount of money, because I don't have the tax base to justify writing my own?" etc. So that's what we see and hear from cities and municipalities when they're looking at IoT.
On the flip side of that, you look at healthcare, which is a hot space for IoT. In healthcare, they're all worried about making sure that the security is second to none because of the liabilities they have if the patient information gets disclosed somehow through a hole in their IoT strategy they did with open source code. So they'll build their own. And so those are two [examples].
I think those are two dynamics that you need to consider -- "you" being anybody that's looking at this. It is, number one, what's motivating a movement to the Internet of Things? It's different for those two parties that I just described. And, number two, what are their risks? The liability a municipality has is relatively minimal. The liability that a hospital has if [your] information gets out is almost inconceivable.
I have a brother that's a chief of staff for a hospital, and the insurance policies that they carry insure them seven times the maximum that a customer could sue them for if they have a security breach. Seven times. Because of fines and penalties, on top of the lawsuit. No city or municipality has that kind of insurance policy. They don't need it. The city of Boston, they'll tell you, "Whatever. Go away. Here's $100 grand. Have a nice day." That would be if you really had their feet to the fire…
TT: It sounds like what I'm hearing from you is that it depends on the market whether open source is a seller, or whether it's a...
GL: An inhibitor.
TT: Or a hindrance.
GL:Yes. Absolutely true. There is not a one-size-fits-all. And people will claim there is, but there's not.
— Joe Stanganelli, Contributing Writer, Telco Transformation