1 month ago
3
ARTICLE AD BOX
AI might be reshaping enterprise technology. But it’s worth remembering that enterprise technology is still being reshaped by the cloud. One organisation that has been instrumental in that is Akamai, which began as a content distribution network and now offers cloud and security services, too. James Kretchmar has been part of that journey since joining the company directly from MIT back in 2004. Officially senior vice president and CTO of Akamai’s Cloud Technology Group, Kretchmar joined as an architect and senior engineer in network operations before working across the network management and mapping groups, taking on CTO roles in both Asia-Pacific Japan and the EMEA region/
In this conversation with Tech Monitor, edited for length and clarity, Kretchmar explains how cloud costs are a problem even for Akamai, how to cut through AI hype, and how innovation doesn’t always move in a straight line.
People see the benefits being reaped from AI and think we’ll all lose our jobs tomorrow, says Akamai’s James Kretchmar. The truth is more complicated. (Photo: Akamai)What are Akamai’s customers telling you about their cloud priorities?
Something that we’ve known for a while is that cloud costs are on the rise for companies. I think we were a little surprised to see both the extent of it and the impact; that the cloud costs are actually driving companies to cut back on other key areas, like AI and other areas of innovation, or even critical areas like security and staffing. It’s something that we actually saw for ourselves and undertook a big initiative to get a lot of our applications off of third-party clouds and onto our Akamai cloud.
In the early days of the cloud, it was a very convenient technology for businesses to use. The growth over time has been very hard for companies to control. And part of that is that there’s a lot of lock-in, right? Companies have found themselves in a place where those costs are going up because there’s organic additional use but, also, it’s really hard for them to have a competitive landscape.
You’ve got deep roots in the open-source community. Some companies have reined back their support for open-source projects. Have you detected some unease or realignment in the open-source world at the moment?
In a constrained environment, everybody’s examining their budgets closely. That can be a little bit tricky, but I think the desire for open source is still there. If anything, we can see some areas where it’s increasing. Companies don’t want to get locked into using proprietary things if they could use an open-source solution.
How is the AI hype playing out? Do you see people embark on projects and think, ‘You might want to rethink that and spend those resources elsewhere?’
There’s a lot of great stuff that AI can do. There are some places where companies are very effective with understanding what they want to accomplish and employing AI to accomplish that. But there are other areas where companies just sort of throw darts at the board, and that’s not as effective. So, I would say understand what you’re trying to accomplish. If you can build an ROI around it, that’s great.
For certain use cases, there’s a lack of clarity around how it delivers the full value. And people hype a certain piece of it and say, ‘Well, it can produce X, it can do Y,’ but that’s not the full value story. For other areas, it’s clearer.
Are there any examples of AI projects where Akamai has sat down, thought things through and successfully navigated those kinds of challenges?
We’re certainly looking at a number of different ways to use AI, but we try to be thoughtful about the right tool to deploy in a given context and what its full value really is.
There are a lot of stats out there that talk about AI producing this percentage of the code at a company. Well, okay, that’s a number, but it’s not the whole story, right? What you really have to understand is not how much code it’s producing, but what makes that code successful. What is the type of work that needs to happen around it in terms of review? Is it more? Is it less? What sort of complexity does it add? Is it successful code?
Does the fact that Akamai is not based in Silicon Valley perhaps contribute to you taking a slightly different approach to AI?
We come out of this very thoughtful, academic type background, where we want to think through things and understand them from first principles. And that probably plays a role.
Here’s kind of a funny example from a totally different industry. My mother used to work for a paper company when I was growing up in the 80s and 90s. At the time, home computers were getting more popular. My mother’s company was getting concerned that the advent and growth of desktops would mean a big drop in paper sales, because newspapers and everything would be online. Now, the reality was, not long after, we also all got home printers – and so, paper exploded.
So, the issue is, with these new technologies, there’s a tendency to want to kind of draw a straight line from A to B. And some people say, ‘You know, it might not be that straight line.’ I think that’s where we are with some of these things like AI. There’s something really important there, but maybe some of the immediate straight-line A to B hype is missing the mark.
