NAN133: Beyond the Hype: The Realities of AI Infrastructure
The Everything Feed - All Packet Pushers Pods
October 14, 2026
Eric Chou speaks with Leo Fleskes on the pressing infrastructure challenges surrounding AI resource consumption, including data centers’ massive water and power demands. Together they discuss local community pushback, regulatory constraints, and the economic viability of AI providers.
Speakers Eric Chou, Leo Fleskes
TopicsTechnology
Eric Chou (0:08)
This episode is sponsored by Meter. If your IT team is stuck managing a patchwork of different vendors, Meter fixes that. They deliver one complete integrated network stack, wired, wireless, and cellular. Plus, managed migrations and upgrade credits make switching risk-free. Book a demo today at meter.com/nan.
That is meter.com/n-a-n.
Hello, and welcome to the Network Animation Nerds podcast, where we explore the latest network animation from a practitioner's perspective. I'm your host, Eric Chou, a network engineer who loves everything about network automation. Our guest today is Leo Fleskes, a network animation engineer at DB Sistel. But before we jump into the conclusion that he's here to talk about network animation or source of truth, as important as they are, right? Leo is actually here to talk about AI resource usage and its impact on our society. This is an incredibly timely and important topic, so let's get into it. Welcome to the show, Leo.
Leo Fleskes (1:11)
Thank you, Eric.
Eric Chou (1:12)
Yeah, so can you tell us a little bit about your background? We mentioned a bit about network animation, but how did you came into technology? Have you always been interested in network animation or network engineering? So tell us a little bit about yourself, please.
Leo Fleskes (1:27)
Yeah, I started programming a bit in high school because I thought it was cool that I could make computers do the things that I wanted.
After school, I started applying to the places and I got, in Germany, we have this thing called an apprenticeship, which is called differently in German, of course. But the point is you work at a company for years and you get paid very little money, but the expectation is that you learn. So they don't expect you to provide economic value, they expect that they teach you things. And I did that. At the same time, I did my bachelor's, this kind of special program. And the bachelor's was computer science and the apprenticeship was more like infrastructure systems integration work. So like the natural cross-section of those two things was network automation. Because at the end of the apprenticeship, I went to the networking department and I did like switching and routing and things like that for a couple of years. But at the same time, I had this software engineering background and I thought, hey, I don't want to for the 100th time configure this thing in our data center core. I would like to have it automated. So I wrote a script that did that. And yeah, that kind of kickstarted that.
Eric Chou (2:34)
They're like, amazing. You can actually tell computers to do that for you without interrupting. Making less mistakes and saves time and everything that people love network automation about.
Leo Fleskes (2:47)
Yeah, exactly that. And then I started hanging out on the network to code slack, the public one, because, well, it was where all the cool people that did this kind of stuff hung out. And at some point, Ken Salenza wrote to me if I wanted to have a job there at NTC. So then let me there and allowed me to further that skill set, because that's of course what NTC does, they do network automation. And I got to learn from pretty cool people there.
And that's also where we met, right?
Eric Chou (3:13)
Yeah, I think we met, well, obviously, we're on two different continents. You're in Europe, and I'm in the US. So the time zone makes it a little bit challenging. But whenever there's company all hands and stuff like that.
But what I think is interesting about your background is you start doing the things already. Like you become the role you want to be almost before you actually got the role. So I think that kind of serves as a good advice for people who wants to do something, right? Just become the change you want to be almost.
Leo Fleskes (3:47)
Yeah, yeah, I feel that. I sat down like after work, I was because I was so interested in the stuff to kind of learn about with Nornia back then, Ansible, NetEco, all the kind of beginner things. And then, like I said, I started applying that and that led me to the NTC job. And then a kind of a similar thing happened with the DB job.
DB was a customer of NTC and still is.
And I realized that I really liked the environment, I liked the people and I liked the thought of working for the German Train Company because I think being able to stand, to wake up every day and doing a job where if you do it really well, you actually make life for everyone a little bit better is a lovely idea to me. And that again, because I was doing like a little bit of architecting, that was not my formal role at NTC, but I was getting into it because I was the only German speaking person on the engagement. Then led me to my role now, which is actually called Nepochautomation Architect, which again was kind of a thing of act as first and then get the job because you acted like it already.
49 more minutes of transcript below
Thousands of transcripts fetched by people building searchable podcast archives
Fetch the whole transcript
The demo key returns a sample episode in full, no card needed:
curl -H "x-api-key: pt_demo" \
https://spoken.md/transcripts/1000651996090Markdown with the speakers named, for your notes, your knowledge base, or anything that makes HTTP calls.
From $0.10 per transcript. No subscription. Credits never expire. Prices exclude VAT, added at checkout for EU customers. Not what you expected? Email us within 14 days with 20 or fewer credits used and we refund the pack in full.
Using your own key:
curl -H "x-api-key: YOUR_KEY" \
https://spoken.md/transcripts/1000796355094