S4E11 Glimpse into the Future of Technical Writing with AI with Caity Cronkhite

Zohra: 0:08

Hello listeners, welcome to Inside TechCom with your host, Zohra Mutabanna. In season 4, I hope to bring to you different perspectives and interests that intersect with our field. Let's get started. Hello listeners, welcome to another episode of Inside TechCom with Zohra Mudabana. Today we are going to have wrap-up to our whole AI theme that we've been talking about. It has been disruptive, no doubt. I have Katie Cronkite with me here, who's going to kind of give us her perspective. She's an entrepreneur and it would be really fascinating to see what she's learned and what are some of the insights that she can share with us. With that, katie, welcome to my show. It's great to have you again.

Caity: 0:55

Nice to see you, zohra. Thank you for having me yet again on your show. Again, it's a pleasure.

Zohra: 1:00

It's a pleasure I get to talk to you and we get to dive right into AI. But before that, how about you reintroduce yourself for my?

Caity: 1:09

listeners Okay, sounds great. Well, my name is Caity Cronkhite and I am the CEO and founder of GoodWords. We are a technical writing and communication consulting firm. Our tagline is that we are on a mission to rid the world of bad documentation, so hopefully we're moving the needle on that over here at GoodWords these days. Before I founded my company, I myself was a technical writer for my entire career, so I've worked in high tech for many, many years as a technical writer and now as an entrepreneur, leading my team of tech writers to help support our clients, and so the AI revolution has come for us. It's here. We've been experimenting with it over here on my team, and I'm actually contributing as a subject matter expert to write a chapter of a book on AI for technical communicators as well. So this is something that I've been really diving into this year and I'm really excited to chat about.

Zohra: 2:04

That's awesome. First of all, congratulations on that chapter. I'm sure it's going to be very resourceful for the community and I completely agree with you moving the needle on getting rid of bad documentation. I think I would absolutely support that mission with you, thank you, thank you. So, Caity, here we are. It's been about nine months and it seems like we are living in a very different world. AI has no doubt been disruptive. It has kind of changed my world inside out at so many levels. Now I can go into many use cases where I have benefited from AI mostly, but from your perspective, just tell me, as an entrepreneur in our field, what has changed for you.

Caity: 2:48

Well, it has been quite an adventure over the past nine months indeed. So obviously I'm an entrepreneur in the technical communication space specifically. So when AI appeared on the scene earlier this year, it was just like a fire alarm fire across the entire industry, and I mean not just across our industry but other industries as well. Right Like it disrupts software development and programming and marketing and all of these other industries that kind of touch the jobs that we do.

Caity: 3:20

So we certainly were not alone in being both very intrigued and a little alarmed when it showed up, and this year has really been all about sort of separating truth from science fiction when it comes to actually using AI at work. So everybody when it first appeared were obviously worried about like is this going to take my job?

Caity: 3:42

Is this going to take away my job function from my company? Are they going to replace me with a robot? And I think that those are, and were, and kind of remain, valid questions. AI can do a lot. It is really really impactful and some of its functionality and capabilities particularly around writing with the large language models, are really interesting and really compelling. But it's been in actually using the technology and trying to apply it to what we do at work where you can really start to see, okay, it does work for this, but not for this. You can really kind of see how this can be a helpful tool in what we do, what it is going to help enhance for us at work and what it can't do yet. So hopefully that answers the question, but yes, it has been a really wild ride the last several months.

Zohra: 4:33

Yes, it has definitely been a wild ride and I think it'll continue to be a wild ride in the foreseeable future. The way I see it, the dust has still not settled. We don't have tight regulation around it. We don't know what we mean by ethics in AI. Right, would you agree?

Caity: 4:51

I completely agree. Yes, there are a lot of unanswered questions. We get a lot of questions in my company from our clients about, like how are we using AI? How can they use AI in their work, what are the best applications for it, what are the challenges and limitations of it? There are so many unanswered questions and things that are that are, frankly, you know, real big question marks that I actually would have thought that we would have gotten a little closer to resolution on by now. But you know, the ethics around AI is certainly one of them.

Caity: 5:23

Regulation around AI, particularly for commercial use and at work, is really interesting. How do you use AI to make your work more efficient, but while avoiding issues like plagiarism, for example, or data breaches or security concerns and things like that? So those are still very much open questions. As far as I am aware, those questions are pretty much in the same spot that they were in earlier this year when AI kind of appeared on the scene. I think that we have a lot more clarity around with AI as it stands right now and what the large language models can do. I think we have a lot more clarity on where it can be helpful at work, but we still have a lot of these big, big open questions that we really need some regulators and some experts to kind of weigh in on for how we can best use it in our jobs.

Zohra: 6:20

The unanswered questions. Those are definitely something that we are going to discover, hopefully find resolution to as time goes by. I do want to kind of dig into hey, what have those questions been that have been answered? But one of the things that I didn't mention was, apart from the whole ethical issue, there's been some discussion about the environmental impact about using AI, because we are going to be consuming a lot of water to support these data centers, or whatever they're called, right exactly, and I just saw one article in the last nine months that has touched upon it, and I think it will become an issue, or probably a challenge, that we will have to address as the adoption picks up.

Caity: 7:08

Yeah, totally. I think that is something important to consider and that's been true not just in AI, but in tech for a long time. As people consume more and more computing power, as people want more and more from their lives online and their digital lives at work and in their personal lives, that we are consuming ever greater amounts of electricity and data usage and GPU space and things like that, absolutely yes. With AI, it's definitely going up even further, right.

Zohra: 7:42

Yeah, it's kind of I think it's kind of really bringing it to the forefront. We may have to consider those issues sooner than later. Probably, with all the other resources we were consuming, we were kicking the can down the road. But that's just my take on it. But that remains one of the unanswered questions so far. Yeah, sustainability for sure. Exactly sustainability, now the answered questions or whatever we have discovered so far, the good part, the fun part, and then we'll also kind of get into the challenges. But from your perspective, how has the landscape changed for our field with AI and all the fears that we hold? But I'm just going to push aside the fears for now and talk about the fun part.

Caity: 8:26

Okay. So the fun parts of using AI in our field right now. I think there are a lot of really exciting things coming down the pipeline and a lot of really exciting applications of AI that technical communicators can start to use right now. So one of the things that's been really interesting is I actually went to a software conference a couple of weeks ago and this particular conference was specifically for founders in the B2B SAS so software as a service space and what I found was that documentation tools are really having a moment In the last 30 years, really before AI kind of came on the scene. I'm sure most of us have had this experience and can attest Like documentation tools don't get a lot of love.

Caity: 9:12

There hasn't been a lot of change in the documentation tools environment for the last 20 to 30 years. A lot of big companies are still using platforms and technologies for technical writing and documentation that were invented in the 80s and 90s. Right, that haven't really changed much. I mean, the Ditto Open Toolkit, for example, has been around forever and has stayed relatively static. It's kind of the same, and I have never gone to one of these software conferences before where I've ever heard anybody mention documentation tools as an area of interest, and at this particular conference a couple of weeks ago, I talked to multiple founders who were really interested in figuring out how to apply AI to the problem of writing technical documentation, maintaining technical documentation, helping developers do documentation better and things like that. So I think that that's pretty exciting. I think it has sort of ushered in this new era of attention on our function and on what we do at work as technical communicators. So that was really interesting and really exciting to me.

Caity: 10:20

So I think what we're going to be seeing a lot of in the next couple of years is AI either being built into existing documentation tools to make them more efficient, to help us do things like write outlines or make editing more efficient or things like that, and I think we'll also potentially see some new players kind of in the doc tool space coming onto the horizon. So I am eerily anticipating what they are going to put together and bring to us for consideration here in the next few months. All of that is still pretty new, though. They're very much in the early stages of development, so it'll be really interesting to see kind of how that evolves.

Caity: 11:03

But I think for individual tech writers, as we think about like, how does AI fit into our lives and into our work, I think there are a lot of places in the process where even low hanging fruit sort of public facing AI tools can be really helpful.

Caity: 11:20

So even just using the $20 version of chat, gpt and GPT-4 has been tremendous for things like editing for straps of documents or creating outlines, or doing brainstorming around the types of deliverables you might want to write for your audiences or what information might this particular audience want to know, or applying certain style guides to writing, so you can use chat GPT to, for example, apply the Microsoft manual of style to a piece of writing, for example. I could go on and on about different potential use cases and applications, but I do think that two major themes that I'm seeing are this sort of area of new upcoming tools that I think will be being developed in the next year or two, and, on the process side, there are ways that we, as tech writers, can apply AI to the jobs and the work that we're doing right now without having to wait for those to come on the market.

Zohra: 12:24

I think I would concur with you in terms of the brainstorming, applying style guides. Those are low hanging fruits that we can absolutely start doing with the public tools that are available. There has been a shift at companies that are restricting access, and that's a challenge. That's probably an outcome of the unanswered questions that we have in front of us, and this is they're taking this extreme step until companies figure out how are we going to use AI responsibly?

Caity: 12:53

Great point.

Zohra: 12:54

Yeah, so definitely those are low hanging fruits that we can start off with, Given that these tools are going to be available to us they're in the pipeline. Not focusing as much on the tools, but in terms of the mindset for technical writers that may be resistant to use AI, or let's say they are considering it but not accessing it Do you have any thoughts on that? What sort of a mindset should they be bringing to this whole new concept?

Caity: 13:21

Absolutely so I can say from my own experience on my own team of writers, when AI first appeared on the scene, we had several conversations with them just to hear what they thought about it, hear what their concerns were, what they're excited about, and on my own team people ran the gamut from on one end of the spectrum we had people who were concerned like is this going to take our jobs? Are people going to hire technical writers anymore when they have these tools? And we had people all the way over on the other end of the spectrum who said there's no way this can ever do anything that we do. We're too specialized, we're too educated in this space. There's absolutely no way that AI can encroach on that. And then lots of people in the middle who thought that AI might change the tools that we use or might be incorporated into some of our work or might make us more efficient and effective, but weren't quite sure necessarily what that was all going to look like. So on my own team I've seen that entire spectrum of people and certainly have talked to other technical writers in the industry as well, who fall all along that spectrum. I think that, no matter what, ai is here to stay and AI is absolutely going to impact our industry.

Caity: 14:41

I think that the people who are on that one extreme end of the spectrum, who think that it's not going to affect anything so far, in what I've seen in the developments in AI and even what I've seen in my own conversations with our clients, I don't think that's realistic.

Caity: 14:57

It's coming for us no matter what in some kind of way.

Caity: 15:01

However, in leaning into AI and using it myself and encouraging some members of my team to use it for certain projects, I've gotten really firsthand experience about what it can do right now and what it can't do right now.

Caity: 15:15

So I think the truth is somewhere in that big nebulous middle where AI tools as they stand right now, can certainly make us more effective at our jobs, can certainly reduce the time and some of the low-level kind of writing drudgery that's involved in technical writing right now. I don't think it is at a place where it can replace us entirely, for several reasons, but I think as far as the advice that I would give to people in the technical writing industry about this right now is, just experiment with it. You don't have to use it to try to do your job. You don't have to use it at work, necessarily, but I do think that it is imperative that we at least get comfortable with knowing what it can and can't do, knowing how to apply AI to the work that we're doing each day, because, sooner rather than later, people are going to I believe, that companies are going to insist that we start incorporating it into our work. That's kind of the experience that I've had so far and what I've seen in my own usage of it.

Zohra: 16:23

You know, Katie, there was a selfish reason for having you here and when you said the whole gamut and I knew that you would bring that broad perspective because you run your own company and you are very well known in the community so you have access to other writers. You could definitely bring that deeper and broader perspective on what the pulse is. I read somewhere AI will not replace us, but writers who know AI will, and I have taken that advice very seriously. I've read that over and over again.

Caity: 16:54

I think that's great advice and I think it's true. I mean, I can tell you from my own experience, my own clients, even with all of the outstanding questions about how best to use AI or the security or compliance risks around it, my own clients are asking me and other people in the industry how are you using AI to reduce cost? How are you using AI to make this project go more smoothly? How are you using it to make this work more efficient? People are already asking for it and they want answers to those questions. As I've said, as we've already discussed, there are still a lot of unanswered questions. There are reasons to tread lightly and not go fully into just trying to use AI for everything at work right now, but people are already asking those questions and they are going to want those answers sooner rather than later.

Zohra: 17:43

Yeah, that dovetails into my next question actually quite smoothly. Has productivity gone up with the writers who have used AI on your team?

Caity: 17:52

I think in some ways yes. So I will say I will caveat this and say that if you are interested in using AI for your work, make sure that you have express consent from your employer before you do that.

Caity: 18:07

There are some as you discussed as well, there have been a couple of instances of security and information and data breaches from people who have used AI without being sanctioned at work. So you definitely don't want to get into that state, but with the permission of your employer. I encourage everybody to at least start experimenting with AI in their jobs and in terms of have we seen it make people more efficient? I think there are areas in which it can absolutely make individuals and departments more efficient at their work. Rit large. I would love to see a comprehensive study done on this. I have not seen those kinds of numbers yet so far in the industry.

Caity: 18:53

But, for example, one area where, just anecdotally speaking, in my own experience, that it saves a lot of time can be with, as I said, editing, for example. So instead of sending every single first draft over to an editor or an editing team, you run that through chat Jpt. It does a pretty good job the first time around If you ask it to just do basic copy editing and things like that. So I think it can make editing tremendously more efficient. I think for certain types of projects, it can have enormous benefits. So one of the areas that I'm excited to apply it to. In some of the work we do is we do a lot of documentation conversion projects, for example, for clients. So an example might be that they have all their content in a knowledge base but they want to put it into Dita format for example, you can literally take an article, copy and paste it into chat Jpt, jpt.

Caity: 19:52

Tell it to apply Dita tagging to it and it will do it in 30 seconds. It's a pretty amazing application of AI for.

Zohra: 20:03

And how long did it take prior to chat Jpt.

Caity: 20:07

Oh, manually. It could take anywhere from five to 10 minutes or more per topic, depending on the state that the content was in. I mean, we've done content conversion projects before that have taken months to do and it's a lot of that sort of manual sort of drudgery of copying and pasting and editing and cutting content down and then applying all the tags to it and making sure it builds and all of that stuff. So for certain applications there are huge, huge time saving implications for applying AI at work. Generally speaking, for individuals using AI, it does have its limitations. Like if you tell chat Jpt to write your documentation for you on a technical product that the world has never seen, it's not going to do a good job. It's going to make up content. It's not going to know about the features that you're specifically describing. It can't talk about things that it has never seen or that the models have not been trained on. So there are things that it is not good for, at least at this time. So I think as long as you have a sort of very specific use case in mind for what to apply AI to, that AI can handle and can handle well, the increases in efficiency can be tremendous.

Caity: 21:32

Like I said, for those content conversion projects, I've never seen anything like it, to be honest. But there are other areas where sometimes it's more trouble than it's worth. Every once in a while I will try to use, just for the sake of experimentation, use chat GBT to write a piece of technical content that I have to work on, and I'll just see how close it can get me, and sometimes I end up fiddling with the prompts for 45 minutes before I abandon ship and it's just like this is not going to work. So I think being really diligent about that, about knowing here's what it can do, here's what it can do well, here's where it can save me time versus, like I can try to use it for this, but is it actually saving anybody anytime? Like maybe not? I think those are important questions to ask.

Zohra: 22:18

Excellent examples and some real world actually with you know we talk about brainstorming, ideation, editing. You're actually giving us solid examples that your team has tried and where they've benefited and where they haven't, and that did I. Example that you gave me that was wow, 30 seconds. I haven't used it up, but I have heard from other writers the frustration that it brings when you're trying to get the tags in and you're not trying to create that magic. So there are definitely, like you said, being diligent, I think is super important about prompting.

Caity: 22:55

And I think that just comes, like we were talking about earlier, from the experience of experimenting with it. I think it's really easy for writers who've never looked at it, never tried it, never touched it, to sort of imagine what it might be able to do or what it maybe can't do. But until you actually start using it and sort of applying it in these real world scenarios, I think that there's no substitute for that right now. And that sort of knowledge about where is it going to help me and where is it not comes from experimenting with it, getting more comfortable using the tool, seeing what it's capable of, what it's not.

Zohra: 23:33

Awesome advice. It really grounds you in how to approach this new technology, technology, new development, and people claim that AI has been around. It's just that we haven't interacted with it at this level, so we've been given much more power with the tools that we have available now, and that's where diligence becomes super important.

Caity: 23:54

Yeah, it's true, I think of it for myself. The sort of metaphor I always use is how people must have felt when the printing press appeared, like I'm sure every scribe on the planet was like oh, there goes my job. But they just had to adapt to learn a new tool, to be able to apply their craft right, and I think that AI is very much like that for us.

Zohra: 24:18

Yeah, I would agree with you. I really think in these nine months I started with that same fear am I going to be replaced? And I could either wallow in that and not move the needle or, like you said, start experimenting. So I've chosen to experiment, and talking to interesting people such as you, it gives me more ideas. Where else could I apply AI? This definitely is very insightful to me personally. One question that I had was from all the examples that you were giving me you talked about, you know you take content and you plonk it into chat GPD. Are you concerned about IP? You know when you put content that's not published, that's not part of the public domain, what are some of the steps that your team takes or that you take to protect that content if you're using chat GPD?

Caity: 25:04

If you are going to take the risk to use chat, gpd or other sort of publicly available products like this, I think you assume somewhat of a risk, no matter what, with that. For example, you know, I don't know if you've heard about the Samsung breach yes, everybody has heard about it. Yes, exactly. So someone from Samsung plonked a bunch of market ideas into chat GPD and, lo and behold, they were leaked to the public there. So I think a few things that you can do, or that companies can do and consider, is obviously, do your due diligence with whatever AI tool you're going to use.

Caity: 25:39

We talk a lot about chat GPD. Chat GPD is sort of the default, but there are others, there are other models, there's barred by Google and things like that. So, kind of as a first line of defense, just like understand what protections you have when you're using each of these tools. It'll be tremendously boring, but read through the privacy policies, you know. Look through some of those user policies that describe you know what's happening with your data and what the risks might be and what you're protected against. So that's definitely something I would recommend like if your company is considering using it, if you as an individual are considering using it at work.

Caity: 26:18

A couple of things I think you can do. You're generally safer if you're using paid versions of these products. Those licenses are a little more all encompassing and protect you a little more. But also, like I said earlier, make sure you're getting express written permission from your employer to use AI at work. You want to make sure that everything is above board, that you are protected in the event of some sort of data event like that and knock on wood, hopefully it'll never happen.

Zohra: 26:49

Knock on wood.

Caity: 26:50

That's a sort of best practice that I've been recommending to everybody is it can be pretty easy to just be like, oh, I'm just going to mess around with chat GPT and like see what I can do with it, but there are implications, you know, if there is a data breach or something like that, that could impact you and your job as an individual there.

Caity: 27:08

So, and I think also another thing to really keep in mind is that plagiarism is a major, major concern. Still. I have used chat GPT myself in my work and in my job and when I go and try to validate the information that it is sending back to me, sometimes I find that it wholesale copied and pasted an article from the internet by a person and that if I were to use that content, I would simply be plagiarizing right. So you want to make sure that you are verifying the information that it gives you, that you are seeking where those sources came from for that information. If you're going to use it in any kind of externally facing document or deliverable or something like that, so that you or your company are not culpable for plagiarism or infringing on anybody's IP and rights there. That really still lies with the individual to make sure that that you are doing your due diligence there and not violating anybody's IP.

Zohra: 28:08

Awesome advice. Is there anything more to add on the fun side?

Caity: 28:12

Oh, I mean, I think AI is really fun in general and I think we're still in for a wild ride. I think it's changing a lot. As I said, I think people are just now starting to really think about applications for how they can incorporate it into tools or into our process at work to really really make a difference. I mean, right now, I would say, most of the technical writers I know are not touching AI for work, and I think that's going to change dramatically in the next year to two years.

Zohra: 28:43

It definitely is. I think I can definitely say that there are some experiments underway. I cannot reveal too much about it, but there are some good experiments that are underway at my place of work that could definitely probably integrate AI into our day to day. I am personally very excited if that happens. But I think diligence is something that we all have to be mindful about and I think that's where companies are starting to kind of tighten their reins and the access that we had just a few weeks ago. We don't have that and that's probably speaking to the Wild West that AI is at this point in time. But having fun is important. While having fun, it is extremely important that we are diligent, we are mindful and everything that you said.

Zohra: 29:29

If you're using it at work, get your employer's consent and do it above board Very important Right now. If you're writing a book, you don't want to upload that book to chat. You're releasing that into the public domain. I have used cloud. I've started using cloud a little bit. I really enjoy it. I use bar. The reason I use bar is if I want to get any sources verified, because chat GPT just lies outright. Cloud accepts that it does not know. Bar provides me with sources. That's just something personally that I've tried and that's how I end up using these three AI tools.

Caity: 30:12

Yeah, totally. And I mean I do appreciate too, with these sort of later releases and updates for all of these AI tools, that they give you a little more control about what kind of content you want to receive. Do you want it to stay factual, Do you want it to offer opinions or different points of view and things like that? So you can sort of tailor that now with the later releases of some of these AI's as well. Very true, which is definitely helpful for me personally and has been for our team at work.

Zohra: 30:40

Yeah, and I think we need to watch ourselves too. We want to protect ourselves. If you're writing something on our own, we still want to protect ourselves. So that diligence and being doing it right for your own good, I think is absolutely important. Now, slightly moving the needle to the more I wouldn't say non-adventurous part. I'm sure you mentioned your clients are starting to ask about the AI tools that your team is using. Are clients asking about your team upping their AI, like getting training, getting on-boarded with AI training, or any such expectations? Are you seeing any such asks, basically from your clients, I think indirectly.

Caity: 31:20

The asks that we hear from our clients and the questions that I hear from executives and leaders, particularly in the tech industry, are that they want and, I think, expect in the next couple of years that AI will meaningfully reduce the cost of certain types of work.

Caity: 31:41

We're already seeing that impact tremendously on AI in marketing content, for example, so it's made a dramatic impact on that industry.

Caity: 31:52

What I'm seeing in my role as someone who is selling technical writing to a number of different clients is they want to know and they want someone to be able to demonstrate, how can AI improve the sort of cost-benefit analysis of investing in documentation or technical writing.

Caity: 32:14

How can they use AI or how are my team using AI to reduce costs to the customer? I've heard about this in larger enterprise technology companies as well that have big tech writing teams. Is they are really meaningfully looking at what are the tools that AI offers right now that can help them be dramatically more efficient in their jobs, can help them maybe do more with fewer people on staff or reduce the time to delivery? So those are the questions they're asking for, but the way we get to that, the way we get to actually reduce the cost of investing in technical writing. For some of these companies is to have people who know how to use it. So I think sort of inadvertently, by asking for and pushing for AI to make some of this work cheaper or more efficient or reduce the timeline for delivering some of these things.

Caity: 33:16

I think on the other side of that coin, what they're asking is like we need people who know how to do this. We need people who understand these tools and understand how to apply that to do this work more effectively. So they're not directly asking the question of like, how are you training people or what training do people need, but what they're asking about is the cost and the efficiency and the timeline for delivery.

Zohra: 33:41

I'm not surprised. I think that probably would have been the first question that any company would ask the cost benefit analysis. That's where they would start With this trend, even though it is indirect. Have you heard from clients who say we would like to reduce our staff because now that AI is available? Have you heard that indirectly as well, by any chance?

Caity: 34:04

I haven't heard it directly, I would say that, yes, I have heard it indirectly. I mean particularly this year. So I'm sure this is the technical communications podcast. I'm sure a lot of people in your audience work in the technology industry right Like so 2023 has been a really interesting year economically to be working in the tech industry. We saw the layoffs earlier this year.

Caity: 34:28

Slow business cycles in tech all year long. I read an article a month or two ago that said that the software industry is down 40% overall from where it was last year, so this has been a sort of remarkable year in the technology industry. We haven't seen these years very often in the last couple of decades, but because of that I saw a lot of this at this conference that I attended as well. People are really really mindful and getting really creative with how they can do more work with less budget, and AI has been a huge, huge factor in that. So people are really really chomping at the bit to get their hands on AI tools that meaningfully make work more efficient and more effective.

Caity: 35:16

I don't know that there are very many examples of actual tools that have been built around functions like technical communication that are ready for prime time. Yet, like I said, I think they're coming, but people are waiting with bated breath for these tools that can actually meaningfully kind of change the amount of time, particularly that some of these tasks take, or the amount of manpower that they take. I haven't really heard from any companies or clients so far who have been able, for example, to replace anyone on their staff with AI at least certainly not in the technical communication space, in marketing and in some of those other functions seen a little bit more of that, but haven't seen it so much in techcom yet. But I think people are watching it. I think people want to see. You know it's like do we really have to hire 10 more people or can we hire five and empower them with some AI tools and get better output from everyone? That's definitely a consideration that is happening right now.

Zohra: 36:22

That's happening right now. Given this scenario, we don't know how this is going to play out. We might have a future where companies retain all 10 employees and upscale them, but there is a 50% likelihood that they might downsize the team to five. In that scenario, I know you touched upon a lot of things that we as individual writers can do, but given this specific play out that can happen in our future, is there any advice that writers should heed to right off the bat? As soon as they hear my episode, they're like okay, I'm going to get down to this.

Caity: 36:57

Yeah, I mean, I think, as we've already discussed, I think, not being afraid of it or be afraid of it, but do it anyway. You know what I mean. Ai is here and AI is going to impact our work. It's going to be built into our tools. It's only a matter of time. Ai is not going to replace writers, but writers who know AI are going to replace writers. I believe that is true indeed, and the thing is is by the time that comes around, there are going to be people in the industry who have leaned in, who have experimented with these products, who understand how to write good prompts and who understand where in their workflow they can apply AI to make it more efficient. You don't want to be on the back end of that learning curve when there are people who have a year or two of experience by the time that comes around. So what I would really recommend now is, like it's free to sign up for an account with chat, gpt, for example, or you can choose another tool, whether it's Bard or Claude or whatever you know, whatever you're comfortable using, but just go apply for a free account, check it out, see what it's about. There are several good, very accessible training programs, a lot of which that are free Right now. I know Google has one for using AI and kind of interacting with AI and writing AI prompts and things like that. That's come very highly recommended. There are ways to get exposure and experience with it that are pretty accessible low, low cost or no cost, that kind of just give you that exposure to it so that as people start bringing this more into the conversation around techcom and doing business in general, at least you'll know what they're talking about Right.

Caity: 38:44

What I've seen a lot which I think maybe you were alluding to as well is I've seen several writers in my own circles who do a lot of speculating about what AI might be able to do but have been too scared to actually lean in and try it.

Caity: 39:00

Maybe they might not admit that that's what's going on, but they've been very avoidant to actually look into these tools or try them out or consider how they might be able to change how they do their work. On one hand, I totally understand where those attitudes come from. This is a big scary change. It's been a big scary change for everybody in the entire industry all year long, but continuing to avoid it isn't going to make it less scary? It's not going to make it. It's not going to change the reality that we're in at this point. So I think earlier this year it was easy to speculate will it or will it not affect our work? And I am willing to go on the record and say confidently that it will. It will change our work at some point. I don't know when exactly that's happening, but it's already happening and it's going to continue to.

Zohra: 39:51

Yeah, only if we had a crystal ball we could look in and say, ok, I'm going to speculate at this point and then start working much later. That's not going to happen. I'm going to ask a question which may seem a little provocative, probably, but or maybe biased, but I'm really trying to get your pulse on it. It's not meant to be a gist or sexist, but have you seen resistant from a certain gender, that is resistant to using AI and certain demographics? That is, within our community? Now that you've interacted with a large community of writers, I'm just curious. I think it's important for us to surface that so that we can start becoming more aware and addressing that early on. That's why I said it might be a little.

Caity: 40:34

In my experience actually, no, I've known some people who have been in this industry for 40 years, who were some of the first adopters of AI tools that I know of. I also know people who are in similar demographics, who have been working in this career for decades, who are like no, I'm not touching it. But I've also seen it on the other end of the spectrum as well, where I've actually been surprised that some of the most technically savvy, technically adept and skilled people who work on the most high tech products and things like that a lot of those people are some of the people who've not touched it or seem resistant to sort of trying it out, which has been very interesting. So I think it's really been all over the place and I haven't necessarily noticed a sort of theme across demographics there.

Caity: 41:24

One thing I have noticed, though, that I think is really interesting and that I think we need to be really careful about. I think certain people, maybe, who have a certain depth of technical skill or something like that, maybe, are resting, have rested in their laurels a little bit. So, for example, I've heard multiple writers who are very, very technical writers say like, oh well, I work on the most technical things, I'm an API writer, or I'm a this or I'm a that. There's no way it's going to do what I can do, so I'm not worried, and I think that's a mistake.

Caity: 41:58

If that's an attitude that you have, I think you really need to examine it, and one of the reasons why I see that, for example, is I was talking to a technical writer who has spent their entire career writing very technical API documentation developer tutorials, developer documentation and things like that and they were very confident that AI could never touch the work that they did, and I frankly disagree. I think API documentation is going to be one of the easier deliverables to automate, with AI probably looking forward into what's coming in documentation tools and what AI can do. Ai can already write tremendous code samples, for example, and just about any language you can imagine. So I think that maybe that is the example of that that I've seen most starkly is certain people with certain skills being what I would consider to be a little overly confident that AI is not going to work, and I would encourage everybody to not take any of that for granted.

Caity: 43:03

I don't think we know exactly where it's going to make the most impact, but I mean AI can write code, AI can write poetry. It's going to impact all of us. It is going to impact all of us.

Zohra: 43:13

And thank you for answering that question. That question was something that I did not prepare for, but as we got through this conversation, it's just something that came to my mind.

Caity: 43:24

Yeah, it would be interesting maybe to see a more comprehensive study about the actual usage of some of these platforms. But, anecdotally speaking, from the people that I work with and the people that I've talked to, it's really individualized and I think it really has more to do with their sort of ideological take on the whole thing than it does about the generation that they belong to or their gender or age or anything like that. But it has been fascinating to see just the wide range of responses.

Zohra: 43:56

Yeah, I think. For me personally, the whole overconfidence thing is a stark reminder that you cannot take this for granted. You got to stay ahead of it because before you know it, it might just displace you and that can be scary.

Caity: 44:12

If we look at the trajectory of how these models were developed, that is the best example that we have. A few months before chatGPT appeared on the scene in early 2023, even just a few months before that, it was still spitting out garbage. Mostly, the models were not trained to be able to sufficiently sort of mimic human language or human output, and over the course of a few weeks, the models developed to such a point where it was pretty nearly indistinguishable. This can and does move really quickly and it will continue to. I've covered this whole gamut of questions. Is there anything you want to add.

Caity: 44:58

I think it's also worth touching on what AI cannot yet do when we consider it in the jobs as technical communicators. So, obviously, it can write outlines, it can edit content for you, it can apply style guides, it can apply certain formatting or tagging to content, but there are a few things where AI cannot do what we do, at least not yet. So some of those examples are as technical writers, we tend to write about new products, new technologies, things that are hot and fresh off the presses, all of the models and the way the models work, large language models that are on the market today. They are generative, so they're trained on these massive, massive corpus of information. Check GPT, for example, is trained on all of the information on the internet up to 2021. But it's really backwards looking. It's trained on this content that already exists and it can only write or produce ideas based on that which already exists, that was written by human.

Caity: 46:09

As technical writers, we're most of the time working on new things that are in development that don't exist in the marketplace yet. So Chat, gpt and all of these AI models cannot yet do that. They cannot write about what is coming and what is future working. That's one really important piece, and I actually what I see and what I predict that's coming for technical communicators as AI becomes more and more prominent is that I think our roles are going to change, but I think they're going to become possibly even more important in that we are going to be the people writing about the new stuff that is going to be then used to train the models, sort of retroactively speaking. So I think that's going to be a huge trend coming up.

Caity: 46:56

Another thing, too, is that these large language models are no replacement for a comprehensive content strategy or for content management systems, for example. So you can use AI or a large language model to write a draft or even write a whole document if you want, but that says nothing about how that document fits into the larger content strategy. Does it help you meet a goal? Does it actually do something that your company needs? A human still has to determine those things, and then from there, where does that content go? Where is it stored? How is it maintained? How does it get updated? Ai cannot yet do those things. There's still a massive need for technical writers and their skills and their sort of like industry standards and best practices that we're specialists in. That content management piece is still there, as far as I can tell, does not change at all with AI. So those are a couple of examples that I think are worth pointing out that even if it can write content, sometimes it can't do everything else that is involved with this role.

Zohra: 48:11

I'm so glad you offered to answer this question and present it to us. I'm kind of flipping that and thinking okay, I'm a technical communicator, how do I stay in the game? And I'm starting to think about content strategy, content management. These are tools that, no matter how much AI progresses and develops, it will always stay behind that. These few things that you've touched upon I think that's what I'm predicting as well these could become your strengths that you could sell to your prospective employer.

Caity: 48:45

And I think they're going to be as important as they've ever been, if not more so, and they're not more so I mean, we see this already One of the biggest challenges in technical communication is not writing the stuff down, it is managing the information so that it's not overwhelming, so that it's searchable, so that it's up to date. We see this all the time with a lot of our clients. For example is like they come to us and they're like we have all this documentation.

Caity: 49:14

We work so hard on it, and then when I ask how it's working for them, they're like oh, oh, it's not, like it's all out of date. Nobody knows how to manage it. Like nobody can find anything. All of those problems have existed for decades and will continue to exist and might even get worse in the short term with AI. I mean, one of the things that AI does is it empowers pretty much anybody out there to write whatever content they want. I think what we've already seen, particularly in like marketing content, for example, is this just like absolute proliferation of AI written content, and that's all well and good, but that just means there's more noise to cut through. Yes, that's the information that really matters. So being able to having those skills in your arsenal of tools is going to be really important. Writing down content does not replace the sort of strategic thinking behind the content, and AI cannot do that.

Zohra: 50:21

Beautifully articulated. I was thinking about content governance, content audit all these become, they just start surfacing. These are either strategic level that you don't think about, that you don't talk about, but that's what we do day to day. It's not just writing that we are, it's not just producing content, but there is all these behind the scenes skills that we bring to the table that we don't talk about. And with the question that you presented to us, you really kind of brought all these to the surface, all the soft skills that we hold but we don't talk about and we don't sell enough.

Caity: 50:55

They're hard skills too. They are hard skills too, actually. Yeah, that's true, I mean, it's totally true. And one thing that I talk to my clients about all the time is that information governance piece Writing the content is maybe 40% of the work.

Caity: 51:14

Most of the work lies in being able to manage the content that you end up with in the short and the long term so that it works for your company and for your users. Ai can do a lot in terms of writing, drafting, editing the content, but then there's this whole 60% of the pie that it doesn't touch. It doesn't really help with at least not now. As writers, even in our title, most of the time, our titles are usually technical writer, but we get very caught up in the actual writing of the words and we sort of forget about this whole other skill set that is required for this function to help companies be successful. And so I think, really leaning into those other skill sets, these other hard skills that are necessary to make your technical communications be successful, I think that that's a really good point. Awesome, I love that.

Zohra: 52:10

I love that and on that note, katie, this has been such a fantastic conversation, so fun, thank you, thank you. Thank you for giving me another chance and coming on my show and educating us. Thank you for giving us another chance. It's been really fun. I enjoyed this. Thank you so much, thank you. Subscribe to the podcast on your favorite app, such as Google, apple or Spotify. For the latest on my show, follow me on LinkedIn or visit me at wwwinsightechcomshow. Catch you on another episode.