Shownotes

Jamie Dobson and Anne Currie

Join us in this enlightening episode as Anne Currie chats with Jamie Dobson, co-founder of Container Solutions and author of Rebels, Visionaries, and Machines. They dive deep into the history of technology, the evolution of the tech industry, and the challenges and opportunities in communicating complex ideas in tech today. This conversation offers valuable insights for tech enthusiasts, educators, and communicators alike.

Key Topics Covered:

• The history of cloud computing and its cultural impact since 1799

• How tech conferences like Software Circus aimed to reshape industry conversations

• The ethical and social implications of AI and automation in the workforce

• The importance of understanding problem context before choosing tools

• Strategies for explaining complex technical ideas without oversimplification

• Lessons from Mary Shelley's Frankenstein applied to tech ethics and responsibility

• The challenge of capturing audience attention in storytelling and education

• How to communicate the essence of microservices, patterns, and cloud-native concepts effectively

• Reflecting on the role of stories and analogies in making tech more accessible

Transcript

Anne Currie (00:04)

So hello and welcome to asynchronous and unreliable, a new weekly podcast where we discuss the most interesting ideas in tech and how they might help us in our jobs. I'm your host Anne Currie co-author of Building Green Software, the Cloud Native Attitude and author of the Science Fiction Panopticon series. I also offer consultancy, training and workshops on getting your business to adopt modern software engineering practices at https://strategically.green.  And for my guest today, who could be more appropriate than my co-author and the instigator of the Cloud Native Attitude book, my friend and long-term collaborator, Jamie Dobson. So Jamie, do you want to introduce yourself?   

Jamie (00:44)

Hello, and thank you for having me on. Yes, I will introduce myself. My name is Jamie Dobson. I am the co-founder of a company called Container Solutions. I always used to say we were the nice people of cloud. Most recently, I've been working on a creative project, Visionaries, Rebels and Machines, which is a history of technology. I'm trying to answer the question, what is the cloud? I had to go all the way back to 1799 to try to answer that question. And I'm just very happy to be speaking with you today.   

Anne Currie (01:13)

It's lovely to have you.  It's great. You represent loads of the things that I want out of this podcast series.  So we've known each other for about 10 years. We met at a really cutting edge conference that you organized a decade ago now called Software Circus that married together all the tech ideas, plus a whole load of stuff about writing and creativity and communications and fun.  It was a really fun event. That was one of the first conferences I'd ever been to and I got there for free because you ran a diversity programme, so my ticket to go was free and I was there. And we first met there and I met your co-founder, Pini, there at that event. And the next year I was the opening keynote for that conference (laughs).   

Jamie (02:01)

Congratulations. I remember we tried to do two things with software circus, make it slightly anti-conference. Back then, all the big conferences in distributed computing were corporate, brushed steel type glass things.  We tried to do something slightly different. And what we did, Mark Coleman and I, was try to coach the speakers on storytelling.  Have a beginning, a middle, and an end. Don't get lost in endless demonstrations, but try to make a point. And of course, we tried to bring more women and people of color into our safe space. But I do remember getting a scathing email from a member of the public saying they would not come to our event because it was free for some women.  And they told us that sounds to me like ladies night at the nightclub. So I'm not going. And I was like, this is not what we were trying to do.  But OK, you can't do right for doing wrong. And that was software circus.   

Anne Currie (02:56)

Yeah, it's interesting how much things have changed over the past decade in terms of what you're allowed to have at conferences. It was quite, it was a bit of an old school men's club. Charles Humble was on the episode earlier, just on the other week. And he was saying about how much he used to get loads of stick for organizing things that said anything about ethics or green or sustainability or anything like that, that we have become much more aware of what's going on in the outside world in the tech industry over the past 10 years.   

Jamie (03:38)

I think a lot of people have really tried. It was about that time I went to see a talk. It was at Go2 in Amsterdam and the speaker had used women in bikinis riding horses.  It was a betting company. And I said to him, what about those images? He said, did you like them?  And I was like, what is the matter with you? What is the matter with you?  And I don't even consider myself that progressive. And I'm like, wow. So yes, things have changed. Sometimes for the better, sometimes for the worst, we seem to take a couple of steps forward and then one step back. The latest round of changes in technology, it's the AI, what do they call it? An AI washing, changes have been made to workforces and people are saying it's because AI doing the jobs. I think we all know that AI has been used as an excuse to terrify workers into keeping quiet. And that's never good for progressive ideas.   

Anne Currie (04:32)

Yeah, it's true. It's being misused at the moment. But at the same time, I am hearing some pretty crazy stuff about how much code folk are being able to produce at high quality.  I've got another podcast that's just about to come out on that subject. So it's both true and not true.  People are being fired. Companies do well, they get a share price boost, for laying a whole load of folk off.   

Jamie (04:57)

We've been trying to figure out the true facts at container solutions for six months. I've been trying to find use cases for generative AI that actually move the needle rather than just help me find references in an old bit of text. I've yet to find any genuinely move needle moving use cases.   

Anne Currie (05:14)

Well, I will say you need to listen to another podcast, which we'll put out more or less the same time as this. So now you've made me think I'll put Martin Davidson out and then you out one after the other.  And then you can go back and listen to the Martin Davidson one. His stuff is mostly experimental. It's not production, but he's someone I've worked with on and off for 30 years and doing really high end, more than production ready, software and he's saying "I've never seen anything like it". He's really using it for high end stuff and it might be that actually it just jumps over all the middle enterprise stuff and just goes straight to the high end stuff which is really expensive to produce and you don't really notice it.  People don't see the value because hardware makes software a lot more efficient. Hardware improvements hide software inefficiency. So people haven't really noticed how much inefficient software there is out there. AI could really, it does seem to be doing very well at rewriting code to be more efficient.  

Jamie (06:30)

My question would be, is your friend a good engineer, a good programmer? So this is what I'm finding when I try to use generative AI or tools based on gen AI to do something. I'm not good at design video editing, I'm utterly lost. So the idea that an idiot, essentially an idiot in my case can team up with an AI to do video editing is bonkers. However, I do know a little bit about writing, but then AI is not very good at that. So I think if it's good at coding.   

Anne Currie (06:34)

Yeah.

Jamie (07:00)

If it's good at coding and you're a good coder together, you'll go really far. Back in real life away from the podcast sphere and all the bullshit on LinkedIn, our customers.  And it's quite confrontational. Are finding that the most effective engineers have become much more effective and those—hate the term trying to think of a better term—less effective or less.   

I'm trying to be polite, not quite as good as the others. Mediocre engineer, less experienced, but basically any mediocrity is also amplified. So companies are out there thinking, well, holy shit, we've got 100 people, 12 of them where they are, are kind of doing everything.  What's the point in having the other 90? Now, I think my question would be, is that true?  Are they really doing all the work? And I guess if it is, well, what are you going to do with the other 90?   

Anne Currie (07:27)

Interestingly, I did say you need to listen to Martin's podcast, which I will cue up before this one, because he specifically talks about exactly the same thing, exactly the same thing. But yeah, it's an amplifier for people who are really good, but then it's probably only an amplifier for them for a while. And then they won't be needed either.   

Jamie (08:03)

Right.   

Anne Currie (08:25)

I mean, he was making the point, which I think that you are as well, which is what's in it for the enterprise to make these changes?  Because all that it leads to is your demise. For everyone, managers, developers, not just developers, it's managers as well.  So it's not gonna be fast. I mean, we've seen this. 

So, rolling back, one of the first things that we did together was you hired me. That was the first time we ran across one another. Effectively, you hired me to write a book about the cloud and were people really doing it because it was the early days of the cloud we were talking about it at conferences. And you were a bit, "yeah, is this all just puff? Are people really doing it? Can you go out and find out." and I like that kind of thing. So I went out and I spoke to whole lot of people, a lot of whom are going to be guests on this podcast. And I wrote up what they were doing on all the kind of cutting edge enterprises, what they were doing in the book. And it was a really good exercise. 

I think what I learned and what we learned from that is that it wasn't really about technology, it was about attitude and a healthy attitude towards trial and error, fundamentally. What we also learned in the 10 years since was at the time, there were some companies who were doing amazing stuff, but most companies were not.  10 years later, that is still the case. It's still more or less those first initial customers who are cloud native. I don't think that the cloud has really taken off and cloud native, actually properly adopting the cloud, still hasn't taken off.   

Jamie (10:09)

Fundamentally, the game hasn't changed. Once upon a time when we mechanized the craft industry, workers needed to change their workflow in order to use these machines. Fast forward to the age of cloud computing and the vendors were trying to convince everybody to go to the cloud, but you could only do it if you changed your workflow. And then surprise, surprise, at the end of that process, you were hopelessly addicted to the cloud, which you didn't really need. 

What we're finding for some tasks at Container Solutions is we'd have to completely redesign our workflows to work alongside Claude, or cowork, for example. But then the big question we're all asking is, "but do we need to do it?" And I think that is the difference between good teams and bad teams is you can use this, you can use Kubernetes to set up a simple application, but do you need to? And because a lot of people can't judge the situation, they're going all in on Claude and forgetting that the real point is to go faster, improve your business processes, reduce your cognitive load. And I think what was true 10 years ago is almost certainly still true today with GenAI.   

Anne Currie (11:14)

Yeah, absolutely. It's all down to actually this—that's a common theme that's coming up through this. I'm really enjoying doing this podcast because I get to talk to a whole load of folk all over the place about what's going on, which is the entire purpose of it. And I'm seeing a lot of people going, look, it's even more important that you work out what your problem is that you need to solve, because it solves problems. And it might be a problem you have, but it might not be the problem that is actually—you might have it, but you might not need to solve it.  There are problems and there are problems. There are problems that you actually need to solve and there are problems that you don't need to solve, you can live with. And there are other more important problems for you to solve first. So understand your business is more important than use a particular tool, isn't it?   

Jamie (11:59)

Absolutely. We had this discussion the other day, somebody wanted to bring Claude into their team and we said, what process are you trying to optimize?  They couldn't answer the question. You probably want to answer that question first before you work out which tools could help you get there.   

Anne Currie (12:12)

Yeah. So the last episode with Jon my husband who you've met, that's just going to go out this week, I think, is we talk a lot about—this is something that comes up over and over that I'm talking to folk about at the moment, which is over-provisioning, particularly over-provisioning of resilience, saying, oh, need five nines, when in fact you only need two nines. And what problem are you trying to solve? Because some of these tools that you would use to solve the problems if you really genuinely needed to be five nines, then it would be worth the massive massive massive investment in making yourself five nines. But if you don't, it's a disaster because it's a massive massive investment and it might not be the problem you have.   

Jamie (13:00)

That's usually a psychological thing. Somebody's anxious and they want to change that into a number at work.  There's never a great reason. Anxiety driving your engineering decisions. That's never wise. What I usually do is say, well, let's talk about cost.  And then they get more anxious about the cost of the five nines. So then they give up.   

Anne Currie (13:12)

Yeah. Which they should because it's really expensive. 

Let's talk a little bit about all the things that you're doing. Because you and I both like the same thing, which is to go out and talk to people and find out what's going on. So what are you up to at the moment that's helping you find out what's going on?   

Jamie (13:43)

I'm only speaking to dead people right now. And so that's causing me great problems.  Well, it's causing me problems in my workflow. So I'm speaking this morning with Mary Shelley and her husband, Percy.  What I'm trying to do is I'm trying to mine history presently for the Visionaries, Rebels and Machines project. Mine lessons and things we've done wrong in the past. I'm trying to learn about the ugly past so we can understand our ugly present.  

Jeanette Winterson, one of my favorite writers, said, Frankenstein is a message in a bottle from 200 years ago. Open the bottle and read the message. So my work is currently time consuming. My work is time consuming because I've got the present, generative AI, people making daft mistakes. I've got to then mine the past, find the lessons, apply them to that audience.   

Jamie (14:39)

it's fantastic. It's rewarding, but it's time consuming. And I refuse to produce any copy using generative AI, even if it could do it, which it can't.  I refuse. So I'm trying to create a trail of breadcrumbs so that the reader can come into the story whenever they like—1950s, 2020—and find some lessons that will apply for them today.  History is really screaming at us right now. It's screaming at us. We've got to listen.  

Anne Currie (15:06)

Yeah, absolutely. Frankenstein, brilliant book if you haven't read it. And for me, it's an interesting one as an author because she produced two versions of it.  There was the original version that just crashed and burned. And then she reissued it. She made some changes, made it bit shorter, got rid of quite a lot of the interesting chapters. And re-released it a long time later, like 20 years later, and it was a giant hit. But actually, I would say go back and read the original release because it's a better book, ironically enough.  But yeah, it's a great book. If you haven't read it in the tech industry, it has so many lessons for an awful lot of the stuff that I used to be involved with about tech ethics, AI, you know, what is the right thing to do, ownership. I would say it's all about ownership, about owning what you did and following up on it.  It's almost long-term maintenance. It's a book about long-term maintenance. What do you think?   

Jamie (16:09)

The techno optimists drive me mad. It's a cliché to call somebody the modern Prometheus.  Oppenheimer was the modern Prometheus. He wasn't actually, but that's what they called him.  There's a book called The Unbound Prometheus. The problem with the tech optimists is Prometheus was chained to a rock and his liver was eaten every day.  He was punished for stealing from the gods. It's not good to be the modern Prometheus. So the real title of the book is not Frankenstein, but it is Frankenstein or the modern Prometheus, which is what the character Victor is. And when he defies nature, when he cobbles together a creature from different parts and brings it back to life or animates it, he is then punished severely.  He loses everything, including his life eventually. And so his ambition destroyed him. He was doing something unnatural. That's the real lesson.   

Anne Currie (16:58)

The beautiful thing about the book is that by creating a human and then abandoning it—abandoning that human, failing to maintain his product that he developed—the product then goes on to be his torturer and finally executioner. The monster is both the main character in the book in many senses and the most sympathetic character in the book, and also the torturer and executioner—the hand of God that punishes him for his hubris.   

Jamie (17:42)

I think the real story of Frankenstein, the real monster is Victor Frankenstein. 

Mary Shelley's mother was Mary Wollstonecraft, a famous—they would call her a radical feminist today, but she was somebody who said the middle classes, especially female members of the middle class, were uneducated because, they genuinely believed this, Anne: after the age of 16 beyond basic arithmetic and sewing, if you educated a woman to university levels, she become fat, deranged and homosexual.  They believed that back then. 

Now, Shelley's mum campaigned against that. She wrote in a famous book—the vindication of the rights of women. She writes in that book something along the lines of any person who is uneducated will not face the challenges that bring to them life's lessons and make them virtuous. This copy can almost be pasted into Mary Shelley Frankenstein. When she was writing it, she was annoyed by Lord Byron, whose daughter was back in England.  He never met his daughter. And he was absolutely devastated to discover his baby boy was actually a baby girl, but she became Ada, the Countess of Lovelace and the world's first computer programmer. 

So I believe the creature is partially a personification of uneducated middle-class women. And I believe that the conflict with Byron as she was writing that shows this is what happens when a father abandons their child, doesn't educate them and starves them of parental affection. They become deformed, which unfortunately is what happens to the creature who then wreaks his revenge. So I think the real monster is Frankenstein and the real lesson is there is a price to pay for men and those around them who abandon their children. I really believe that's what it's about.   

Anne Currie (19:37)

Yeah, that's a very plausible explanation. But as I say, the good thing about Frankenstein is it's a rip-roaring read and it's a genuine page turner. It's well written, genuine page turner, but also it's sufficiently full of stuff and ideas that you can interpret it—there's loads of threads of lesson from it, there's loads of stuff and you go, that's an analogy for this, but that's also analogy for this. And actually, it's not—as an author myself and for you as well—it's not untrue that for a good book, you should be able to apply multiple analogies to the same story.   

Jamie (20:31)

Yeah. 

It's not only a great work of science fiction, it's the first work of science fiction arguably—began the genre—and 200 years later, it's still doing what good sci-fi does, which is make the reader stop and think.   

Anne Currie (20:46)

Totally. Yeah, it's actually an enjoyable read. Sometimes when people recommend a book, it's a bit turgid and you kind of think—especially something which is a Penguin classic—you're thinking, that sounds bit worthy to me.  But it's not worthy. Well, it is worthy, but it's actually enjoyable as well.   

So that's your current obsession and that's pretty good. So Visionaries, Rebels and Machines is the book that you wrote, which is very good about the history of the development of technology. Are you now working on a second?   

Jamie (21:28)

Well, so Visionaries, Rebel Machines, I've done it all backwards.  I'm working this out as I go along. So I'm back after being a manager for decades.  I'm now back creating and I love it. I wrote the book. A lot of the work went into the research, trying to build those breadcrumbs from 1799 with Volta.  That is the time of Frankenstein. The battery—Mary Shelley's book was a reaction to the battery. 

Jamie (23:06)

Volta's conflict with Galvani divorced neurology and electronics. But actually, they become reconciled in the innards of a machine called the perceptron, which is essentially the precursor to artificial neural networks. So my story starts in 1799 with the divorce of these technologies, reunites in the 1950s in the perceptron, and then gives us the world we've got today. So what am I doing is I'm writing a substack with lots of breadcrumbs that bring people to the book. And I'm trying—I make the sign of the cross—I'm trying, I don't know if I'm succeeding, to write a single narrator podcast using occasionally silly voices, but screenwriting techniques to try to introduce these stories and characters.  

Some of them have done really well. Some of them have done less well. What I'm discovering is that people's attention spans are just too short. And that's a real challenge for me because how do I explain information theory properly?  It's hard to do in 20 minutes. So how do I balance talking about information theory in a way that explains it, but without boring the audience to death?  And that is my great experiment right now. And I'm proud of myself, despite how insecure I get when I'm doing my stupid voices.  I'm glad I'm trying something. And I do believe there's a story to be told. I think the book does a pretty good job of that, but now I'm trying to broaden the reach and take this visionaries and rebels and machines idea to the rest of the world as it were.   

Anne Currie (24:37)

Yeah, so basically you're trying to devise the Sesame Street version of distributed systems and information flow theory.   

Jamie (24:50)

Every single chapter ends in that book apart from the last chapter, where to now? I am desperately trying to give people the breadcrumbs and so they ask that question where to now? And I promise I can teach the reader and the listener—I can go from a switch that's on or off and take them all the way to neural networks—but you've got to go in step by step.  Now, my biggest challenge is I'm not sure if anybody cares. Now, that's the problem. If nobody actually cares, I'm in big trouble, but assuming at least a few people care, I want to know, do we go from a switch that's on and off to an artificial neural network?  Then I might find my audience. So that's the challenge I've set myself.   

Anne Currie (25:29)

Well, I'll read it or I'll watch it or I'll watch your podcast. And I'm really interested. So, one of the other themes of my podcast is how you communicate difficult subjects. Your first, rolling all the way back 10 years to software circus, and it was the first iteration of, I would say, one of the most significant talks that happened in DevOps ever, which was the first time Kelsey Hightower demoed bin packing for applications. Higher density machine utilization. 

It was pre Kubernetes. He wasn't even working on Kubernetes at the time. And he did it by playing Tetris live on stage.  And was the first time he'd ever done that. He was the closing keynote, I think, for the first day. It's still kind of like—I was in the audience for that. And the next day, oddly enough, I'd been working on the area myself and I went up and I introduced myself to him the next day and said, I really enjoyed your talk.  And he said, I was quite nervous about giving that talk because I'd never done it before. I didn't know how well it would land. And it was an absolute game changer of a talk. 

It was so much better as a way of demonstrating what early bin packing and schedulers and automated programmatic data centers could achieve in terms of increased machine utilization.  Really, really, really dry topics that are hard to engage people with. He did something very visual on stage. It was really astonishing.   

Jamie (27:16)

This is the Holy Grail, isn't it? We want to make concepts explainable, but we don't want to dumb them down so that, at some point, you have to go through the levels. And I think that Kelsey's talk was the starting point for many people.

I think what I'm trying to do with visionaries, rebels and machines is everybody's scared of computers in the same way that people used to be scared of programming their VCRs.  People were scared of machinery. My grandma was afraid of the washing machine. And my mom used to stay off school so she could switch the washing machine on. It was just a switch that said 40 degrees or 60 degrees or whatever it was. 

But we cannot get away from the world of computing. The world of computing and the tentacles of gen AI are right into society now and into our children's lives.  So we can't opt out. Now that's one problem. Another problem is if it becomes so confusing, people just check out. And then that's when the demigods and the populists come in and take advantage of it. 

How do we as technologists—and it's such hard work, every time I get a sentence that explains something concisely, it's usually because I've been working on it for about six months. And so finally the sentence appears and then I've got it and I can take the reader with me. How do we as ethical technologists put these messages in a way that get people started and then lead them onto bigger and better questions so that they genuinely understand the world and can navigate it?  I think that's what I'm trying to figure out. I think that's what you're trying to figure out. And I know it could be done, but it's not easy.   

Anne Currie (28:55)

It's really not easy. So we've both organized conferences.  I've been at your conferences. We have organized conferences together. We've worked together. I was one of the nice people in the cloud at Container Solutions for quite a few years. And yeah, throughout all of that, when we've written books, we've done all we've done. We've been on so many podcasts together, but I still think we're still searching, aren't we, for what is the way to communicate difficult topics?   

Jamie (29:24)

I think if you look at the cloud native attitude, it's a great primer. And so how I describe the cloud native attitude is that there are lots of books on microservices—Sam's books. There's lots of books on observability, but there's no broad brushstroke. What is the cloud starting with continuous delivery, orchestrators, etc. That's what the cloud native attitude does. 

Then you move over to cloud native transformation and the pattern language. That's one level deeper, not technically, but how do you recombine these patterns to succeed with cloud computing or cloud native computing? 

And then, of course, if you want to go one step deeper, if you want to look into microservices, you go to Sam's book. There are too many specialized books, in my opinion.   

Anne Currie (30:06)

So Sam Newman will be another guest on this podcast later on.   

Jamie (30:09)

Good. That's great. Yeah. So there's too many specialized books. And so if you look at the history of computing, there's a great book on Xerox and there's a great book on Edison and there's a great book on Amazon, but who has drawn the bow and shot the arrow right across time.  It needs to be done. Those stories with beginning, middles and ends need to be told. So we have a sort of binding and understanding of what's under that umbrella.   

Anne Currie (30:34)

Yeah, I know that's your key thing that you want to do at the moment. Now, we have been talking for half an hour, and I promised myself I'm going to try and keep everything shorter, and have them more often because I've been going a bit crazy in the last two and I've gone for like over an hour, and then it becomes really hard to edit them.   

Jamie (30:56)

Yeah, absolutely, yeah. Short and sweet is the way.   

Anne Currie (31:01)

But also we've had such a good conversation. I've really enjoyed this and I'm hoping that I'm going to be able to entice you back to talk over and over again on this podcast, Jamie, because I always really enjoy our conversations.  So let's draw this one to a close. We can go on talking after this, but let's draw this one to a close. And then I will say to the viewers, thank you. And the listeners, thank you very much indeed for listening to this episode of asynchronous and unreliable podcast. Vastly would vastly appreciate a subscribe or a thumbs up on YouTube, whatever you like helps motivate us to keep going.  But thank you very much. And thank you very much, Jamie, for being an early guest on the podcast.  I do appreciate it.  

Jamie (31:55)

Thank you.