S05E31 · Des applications plus vertes et le SwitchLightOps

0:03 A problem a lot of people who like making things have, including me, is that sometimes, you know, you spend quite a long time making something and then you realize at the very end that actually you're solving the wrong problem. I get really frustrated by waste. And sometimes the things that we do to try and be efficient don't actually help us be efficient. Seeing something that's a little bit outside my day job allows me to make those connections in my day job and realize, oh, this is the technology I need, or this is the way I need to think about this problem. Hello everyone, I'm Philippe Ensarguet, VP Software Engineering at Orange, but today it's as a member of the Tech.Rocks content team that I'm introducing myself to you. I'm delighted to be with you for this Tech.Rocks podcast for which I have the pleasure of welcoming Oli Cummins. Hello. Oli will be speaking at TechRock Summit 2023, which will be held over two days on 7 and 8 December at the Comet Meeting Booth in Paris.

1:10 This will be the seventh edition of the event, which will be held in hybrid mode and will welcome more than 500 people with CTO, VP and head of technology profiles. It is an important moment because it brings our community together. It's a moment of inspiration, reflection, and a perfect opportunity to network and meet new contacts to strengthen our practices and learn together. Let's start by meeting Holly, better understanding who she is, her background, and the scope of her session. So, Holly, I'm sure our audience is curious to understand who you are and what is your background. So I'm Holly Cummins. I work for Red Hat, but I've worked for Red Hat for about a year and a half, and I've spent most of my career at IBM. I've done quite a few different jobs. I like variety. Right now, I'm a senior principal engineer, and I'm helping to develop Quark. I've had engineering roles in the past, but I've also had some consultant roles in the past where I was working in a team called the IBM Garage.

2:18 What we were really trying to do was help organizations adopt lean, help organizations adopt design thinking and see how development methodologies like extreme programming could really help organizations take advantage of the cloud, which was quite a new technology at the time. And to be honest, actually, I think sometimes the cloud still feels like a new technology. I think we still don't always get the most out of the cloud. Perfect. Thanks, Solly. I would be very curious about what drives you and gets you up in the morning. What are your daily inspirations? I mean, fundamentally, I think there's two sort of things that drive me that are a bit opposed. One is that I just love making things. And always in my career, I've tried to find things where I can make things. And then you hope. That the things that you make will be useful and will make someone's lives better. But then the other thing that really motivates me, I think, is I absolutely hate waste. And so quite a lot of what I've done in my career has been looking at processes and seeing that they just didn't make a lot of sense.

3:21 And that if we would change how we would do something, if we would automate what we would do, then it would be better for everybody. But I mentioned those two things are slightly contradictory because I think a problem a lot of people who like making things have, including me, is that sometimes you spend quite a long time making something and then you spend quite a long time making something. you realize at the very end that actually you were solving the wrong problem and you've made something amazing and you should have just been a bit more tactical and, you know, reused someone else's software or just accepted that even though it was a slight irritation, there were higher priorities. And so sometimes I end up making the wrong things. After scaling last year, the theme of this year is about efficiency. And I think our audience is really keen to hear what that means to you, how it resonates for you. Yeah, I love this theme because it is something that I've sort of struggled with and thought about for a lot of my career. Like I mentioned that I get really frustrated by waste. And sometimes the things that we do to try and be efficient don't actually help us be efficient.

4:28 So for me, for example, I have a bit of a hobby of automating processes. And there's the classic thing that probably everybody has seen where you set out to automate a process because you want to make things more efficient. And two days later, you've forgotten what the original problem you were trying to solve was because you're deep in, you know, you've written a whole framework and you've written a whole set of scripts. And then a month later, you're still working on the automation and everybody around you has forgotten what your day job even was. And so when we always, you know, sort of set out to be efficient with... really good intentions, but often we can get these unintended consequences. And when we think about efficiency as well, there's a couple of different kinds of efficiency that are interconnected. One is the technical efficiency, the performance. How can I make this software more efficient? How can I make it use fewer resources? How can I make it use less energy, which is really important. But then the other aspect is that people efficiency of how can I make my team more effective?

5:30 How can I make sure that my team's time is spent on stuff which is going to add value, add business value, rather than just spinning the wheels? Looking like there's activity, but it not actually going in any useful direction. Interesting, very interesting. So now if we are getting into your session, what are you going to talk about? I'm really going to explore these aspects of efficiency and how the technical efficiency and the people efficiency connect to one another. I'm going to explore some efficiency anti-patterns because, you know, this is something that I think a lot of us have seen that you set out to be efficient and you actually end up. With the opposite and how did this happen? And then more generally as well, as a species, we are pretty terrible at being efficient. And how, after so many years of evolution, do we still end up designing these processes that are horrible for everybody involved? Why didn't we learn the lessons from the last 2000 years of written civilization?

6:35 So that's sort of one side of it. And then the other side of it is the technical efficiency. But then, of course, what we're seeing now is sort of a new intersection of these two as well, because we have AI disrupting everything, really. And I think the question everybody wants to know is, is AI going to make us more efficient? Or are we just going to end up sort of doing the equivalent of watching cat pictures on the Internet with these tools? And so then what does it mean for our jobs? What does it mean for our businesses? I would love to hear you about the kind of outcomes that the participants will be able to bring on. Basically, what will they learn? I'm hoping to sort of introduce a toolkit for thinking about efficiency and thinking about how to make things efficient, thinking about how to spot these patterns where something became inefficient despite our best intentions, and a set of questions to ask as well when we try and make something efficient in terms of either unintended consequences of this thing that's supposed to make us more efficient.

7:43 There is a question I would love to ask you. I'm really wondering how you manage the preparation of your session. Did you have an initial strategy? And perhaps the most important, did your opinion change about the topic during the exercise you made? I can start at the end. actually, because one of the things I do talk about is the impact of tools like ChatGPT. And so at the very end, I wanted to put a slide in showing some example ChatGPT output. So I put in the topic of my talk, the title of my talk. And ChatGPT came out with something that was very similar to what I had already prepared, which was incredibly depressing. But then actually, when I looked at it a little bit more, I think the sort of the classic pattern that we see with these kinds of tools that on the surface, it looked like I had just completely been replaced by ChatGPT. But then when I looked at it more, I realized that what it was saying was quite superficial and that it was actually missing some connections that I had made.

8:44 So that was the sort of the final process that I didn't take, was just feeding the whole talk into ChatGPT. But I find for this kind of talk, what I really like to do is seed the idea and start thinking about it and then leave it for a while. And so we started talking about this talk actually quite early in the spring, in terms of the organization, actually. I know you said, can I give you back an abstract quite early? And then there was a bit of silence and a bit more silence and a bit more silence. And, you know, I hadn't forgotten about it, but I was thinking about it. And I find that actually it's a gain. It sort of relates to efficiency that if you. For something like this, if you have a time box session, you say, right now you're going to have an hour and you're going to come up with thoughts on efficiency. It ends up being fairly superficial. But if you have six. months and you're not really thinking about it but while you're in the shower it's percolating away in the background and while you're going for a walk in the forest is percolating away in the background and actually quite a lot of this talk I wrote while I was watching a Tottenham Hotspur game because again it's just that sort of that it's called the the default mode network

9:49 that when we are doing something else, our brain can really do quite meaningful, important work on a more difficult problem. And so I sort of, it, all of those thoughts percolated away for that time. And then I pulled them back together to come up with a sort of a central thesis. And I think in terms of my opinion, when I started, I knew there was contradictions with efficiency and it's a good thing and a bad thing and it's easy to achieve and it's hard to achieve. And at the end, I think I still have some of those same opinions, so I think there is still some contradictions. Thanks so much, Sholi. I think it's really, really interesting. And it was one of the questions that I've got in my mind since the very beginning. We knew we love you. One final question, because there are still some tickets available. What would you say, I would say in terms of teasing to perhaps make those who haven't got their ticket yet want to join us for the event?

10:49 I think with this kind of event, there's two things that are absolutely brilliant. One is... The talks and what you learn from the talks. And particularly what I find most valuable is the talks that I didn't know I didn't know about. So when sometimes I'll sort of drift into a talk without paying very much attention, and then I'll realize it's something that I wouldn't have chosen to go to. I certainly wouldn't have gone to YouTube and hunted down a video on that topic. But actually, seeing something that's a little bit outside my day job allows me to make those connections in my day job and realize, oh, this is the technology I need, or this is the way I need to think about this. problem. But then the other thing that I really get out of this kind of event is the people you meet. And we call it the hallway track, don't we? That those connections give you ideas, but then they're also something that you take away. And then it's a network that you have that you didn't have before that you can draw on.

11:50 Holly, we are reaching the end of our podcast. Do you want to share perhaps one last thing? Yes, I was going to share this in my talk, but I will give it as a spoiler. Because do you know what the most efficient animal is in terms of movement and conversion of food into movement and that kind of thing? Whale. That probably depends exactly how you define the question, actually, which is, of course, one of the challenges with efficiency. Because the answer that I have is if you look at just land animals, then the most efficient animal is a kangaroo, which kind of intuitively makes a lot of sense, right? They're very bouncy. They go quite fast and they go quite high. But if you look more broadly and include the ocean, which, of course, is where whales are, What I read is that the most efficient animal is a jellyfish. And when you sort of think of your ideal of efficiency, right, you imagine something like, you know, a tiger or, you know, a cheetah, something that, you know, is really sleek and fast and strong. But actually, sometimes efficiency doesn't look like that.

12:51 Sometimes efficiency looks a bit sort of blobby. Yeah, funny and very interesting, if I may say. So just a quick words for me. I was very lucky to be able to attend to your real soul. And I can tell to our audience that I really love the intensity and the conclusion brought that will question definitely ourselves about our real priorities in the face of the challenges of efficiency. Holly, the time spent in your company has gone very, very quickly. It was totally captivating and exciting at the same time. And I'm sure our audience will, of course, be able to interact directly with you at the summit. So thanks so much for being with us for this summit. So in the name of the organization team, we look forward to seeing you all at the Tech.Rocks Summit 2023 on 7 and 8 December at Comet Meeting Booths in Paris. I'm really looking forward to it. Thanks, Oli.