API Evangelist Conversation with Sue Smith, Senior Learning Experience Manager at Fastly

Hi, I’m Sue Smith.

Um, that’s a good question. And my description of it changes frequently. Um, how I’m going to describe it now is that I make learning experiences to help people build with technology. Usually developer technology and usually for the web.

I’ve enjoyed the heavy sigh when you said DevRel there. I felt that. Do you know what? It’s always been the kind of poor relation in the, I’m going to say the age of DevRel. Um, education was kind of included in whatever DevRel was, which varied wildly between companies, didn’t it? But it was always a kind of neglected piece of it. Um, I believe that that’s changing because What DevRel was is radically changing for lots of different reasons. I think partly because startups are changing, because the way that funding works is changing. Um, I don’t think we have the kind of luxury that we had to not be able to demonstrate impact. Effectively through that kind of work, um, I think that we probably focus too much on things like awareness and not enough on enablement. And for me it was interesting over the years to see that people in more customer facing functions, people who are maybe in the sales cycle, started to invest more in education. Then we did in developer community settings like DevRel. So there was some understanding that it was a necessary piece of the puzzle to enable people around a technology product. Um, what that looks like in the future, I’m not sure, but I do believe that education is one of the pieces of what we used to call developer relations that will survive.

You know, the fun piece is interesting, because I think one of the reasons that education wasn’t invested in heavily was that it’s not cool. It’s not like shiny and interesting. But in terms of the fundamentals, I don’t know what those are anymore. I don’t know what the fundamentals are going to be for an AI assisted coding. reality. I don’t know what people are going to need to know. And it’s always been difficult when you, you know, when you teach people about technology, it’s really difficult to keep track of changes in the industry and ecosystem and to be confident that you’re teaching people valuable skills that they’re going to go on to use because it changes so rapidly. But right now it’s, it’s changing so fundamentally that I don’t know what people are going to need to learn and I don’t know what they’re going to be able to just outsource to some tool.

Do you know what, in some ways it is changing fast, but I also think that the AI situation is kind of forcing us to better articulate those things. You know, a few months ago when a lot of these AI assisted coding tools just suddenly exploded and there was this panic that we’re, we’re going to make software engineers obsolete and to me that was a reflection of the fact that we don’t have a compelling, robust definition of what a software engineer does, of what the skill set is. And there seems to be this, in my view, misunderstanding that it’s writing code You know that it’s writing syntax, but it’s not that’s that’s the detail Um, and I thought would be really interesting to me was would be for us to come up with a real Definition of what a software engineer does, you know If you let’s say if you look at a more experienced engineer and you were to articulate what kind of value they bring to an organization You know writing the correct syntax wouldn’t make the list would it? It’s that’s not the skill set It’s a more abstract set of skills You That they’re able to deploy in different settings, different languages, different frameworks, different problems. Um, probably the more API relevant piece of that is, I think, a huge part of software engineering now is managing complexity because we build software in such a modular way, where we’re plugging all these little Lego block components in, and that makes the managing of all that complexity is a really huge challenging part of the job, and I have a suspicion that it might be one of the parts that is least automatable, I feel like that’s one of the pieces we’re going to need a human for because it’s kind of complex and nuanced and it depends on like an understanding of the context, business context, technical context, all the rest of it. In fact, that’s a point where I would like to ask you a question. You mentioned online, I think it was probably on LinkedIn, something about AI and that perhaps necessitating I return to the notion of the semantic API. Am I remembering that right? Yeah. I’d love to know more about that.

Yeah. I mean, uh, the semantic stuff is, It’s one of those cycles in API land that kind of has come and gone several times and, and, and it’s gone away. You don’t hear a lot of folks talking about it. So, uh, JSON LD is a, is a semantic way of expressing, uh, JSON and using it. But that’s part of a larger academic tradition around, uh, or discipline around, uh, language and semantics and ontology and linked data. And so there is actually a lot more, you speak of engineering and kind of discipline, I think semantics has has more of a root in the real world, but it just was never deemed useful enough in, in API circles and JSON LD is actually very used. It’s used if you, uh, schema. org. So if you go to schema. org and, and you, uh, you, Everything’s represented as Jason Ld there. And if you want to zoom out more practical when you google something, you search and you see those little cards at the top. So if you search for Sue Smith, you know, if you search for fastly, if you search For these, these concepts that are a people, a person, a company, these are all, uh, semantically labeled with Jason LD. And that’s why you see that card at the top. So a book is a book, an event is an event. And so these semantics are really important, but they’re not easy and you have to do a lot of work and. I’m definitely not an expert. I’m very new to that space. But every time I dive in and spend time and work with, uh, folks in academia who are doing interesting things there, I realize how much I don’t know and how much foundation that is for me. Language, meaning relationships. There’s a lot that goes into it before you actually get to the actual semantics. There’s a lot more foundational things that you have to go into and so kind of bringing this back to, um, hey, I I’m just not convinced there’s there’s enough will to do the work there because if if we weren’t willing to semantically, uh, enrich and decorate our API’s before, uh, And our data. I just don’t have a lot of faith that we’re going to do it in service of AI and we’re going to expect like AI to do this. And, and there’s a lot of devil in those details. And I don’t think we, back to your word, use of the word engineer. Um, I think we like to play engineer and call ourselves engineers. And I don’t think we are. I mean, if you think about building architectural engineers, you know, that are building things and, and kind of what you were alluding to is. You know, the code itself is just an output, you know, the bridge or the building is, is just the results of that work. But what are all the things that an engineer has to, has to learn, you know, physics, um, you know, geology. There’s a lot of things that have to come in to know how, if that building or bridge is going to stand. And do you feel like those things exist in our world that we have enough discipline and have been doing that?

Absolutely not. Do you know what else I would like from traditional engineering disciplines? Apprenticeships. Wouldn’t that be nice if we didn’t keep the hell out of tech jobs and we actually trained people on the job and you were able to access these jobs without having had access to some sort of super privileged degree? That’s a side rant. I’m not going to go on. But this is really interesting to me because one of the things I’m doing just now Um, Um, I, where I work at the moment at Fastly, we have an edge computing platform. So I’m trying to teach people about edge computing. And a huge part of the challenge is conceptual. And I’m trying to articulate to my co workers that for a technology like that, that you have to leverage abstractions and mental models and that kind of thing. And so I’ve been doing a bit of reading about it, and code comprehension and that kind of thing. And one of the distinctions I came across. Was that when an engineer is understanding a system that the knowledge that they have about that system is both semantic and syntactic. So the semantic knowledge being like understanding how the components interact, how it interacts with the context around it, technical context. You could maybe even stretch that to say. Understanding the business or socio cultural context, even, and it kind of sounds like, if I’m understanding correctly, what you’re talking about, if we did actually invest in that, would be some sort of machine processable version of that knowledge? Does that sound right?

No, I think so. I think, and I think there’s a lot of groundwork laid when it comes to, Networking, uh, when it comes to, uh, the how the internet works, how the internet came to be that, um, people gloss over and people just take for granted at this point. And then you don’t need to understand it. It just works. Um, but how, how, how that large system works. works. And I think this is why we’ve ended up with graph QL and web sockets and other things. Not that there’s not a need for those layers. I’m not trying to bash those, but I just don’t think people have done the work to go. Well, what are we doing on graph on H T T P? What are we doing with the basics of networking? What, uh, what models of of Are there out there? What patterns? There’s a lot of architectural patterns when it comes to software that are 30 40 50 years old. Um, and they’re not even near as old as engineering, classic engineering concepts. They’re still pretty in in their infancy. And I just don’t think we have, um, The affordances, the money, the resources and time to, uh, to teach people these things, uh, in their jobs. It’s, it’s, it’s the underlying funding model, the winds changing. Um, and I don’t think people see the, the negative influence of venture capital in changing the winds. And not prioritizing these foundational concepts and people learning them.

Totally something else i’ve started becoming aware of recently with having lost track I’ve gotten older and haven’t been around in the tech industry for a while is that For a lot of younger people the way that they’ve accessed the web It’s been through platforms through walled gardens and all that and so one of the things that i’m spending about my time doing now is explaining to people some of those web fundamentals like Why it’s a good idea to have your own domain, that kind of thing that people have just never even been introduced to these ideas and they don’t necessarily have a solid distinction between like looking at an app on your phone and the web, you know, then there’s so much there that’s also that’s tight that’s part of the same story because it’s about these big platforms that wanted to Keep people within the bounds of their apps, isn’t it? And I think we have even more work to do now because of that lost all that knowledge

Yeah, it’s interesting because when people started leaving Twitter en masse, it seemed like there was a moment when people started to become aware of this idea of an algorithmic timeline, and that there was something going on. determining your experience and the information you were exposed to that was not transparent and that wasn’t under your control. Um, and platforms like Mastodon, I feel like there was a brief opportunity there to use that to educate people, but unfortunately it suffered from something that a lot of tech suffers from and that it’s extremely difficult. for people to use if they don’t have like 14 advanced degrees in computer science. Uh, I’ve seen one of the things I find really encouraging is that I see young people on TikTok getting angry about the fact that, like, they’re looking at the feed on their phone and their pals seeing something different. And they’re going, wait a minute, what’s going on here? And it’s not just the fact that it’s not a chronological feed, it’s the fact that there’s some mechanism that they can’t see. You know, and that’s a tremendous opportunity for education.

I think so, because especially for that platform in particular, it’s such a kind of empowered community of young people. They want to know, they want to understand the mechanisms of power. They want to know what the dynamics are that are leading to the experiences that they’re having, wherever it is, including the web.

I have a mortgage to pay. Do you know, I’ve never really lost my love for doing this education work because education changed the trajectory of my life. When I was younger, I don’t, I don’t think I had a future. That sounds melodramatic. But, you know, I came from a poor background. It was a rough time. And I had access to this course that was free. Because the Scottish government was funding any software course, and it was a foundation software development, and it just put my life on a completely different trajectory. I’ve had so many opportunities, I’ve had an experience of life that I couldn’t have imagined. And I want other people to have that experience. And there’s also, it’s also slightly fueled by the rage at how gatekept and privileged Terk is. That fuels me to an extent as well.