This interview discusses using AI (ChatGPT) to simplify and standardize internal wiki pages by treating AI as a junior employee—providing it context and then using it to translate expert-level content into simpler language. The process keeps experts central (they provide original content and review AI outputs) while compressing iterations (draft → AI simplification → expert review → beta testing). The team is empowering community managers to prioritize, create, and own wiki pages so knowledge is embedded in teams and updated faster.
Chapters
I think most people assume AI is like some kind of oracle, wizardy, super expert thing. And my view is you should write for AI like it's actually a very junior employee. You should imagine it's an intern and you give it the context you give an intern and the same kind of thing. You're talking about the level at which we should be writing. Alan. But the interesting point is most people that write a knowledge management or they teach learning and development courses or whatever, internal comms, it's usually the experts or at least kind of more senior people or the person that knows the most about the thing writing it, but they're so far away from where that intern or that novice or the AI is that you almost need someone to come back and use or be a tester, you know, to like see if they can comprehend it. Like you need a, a junior level person to read this. Do I understand what the page even says?
You know, but this is exactly what we're doing and this is how we're creating the workflow to create all of these new wiki pages. Right? So AI is not writing these wiki pages from scratch. People are still writing the content. They're still putting their expert thoughts down in what it needs to be. AI is helping me bring it down to the level that anybody can understand it. AI is helping me to simplify the language. So I'm training AI to understand the level of simplicity and clarity that we want consistently across all of our knowledge. So people say, oh, AI is taking jobs away. It's not taking jobs away from me. It's going to allow us to get through this massive database of rewriting these wiki pages way more efficiently. But I still need the experts to review the process because AI does not know what the process is. At Turner Fleischer but if I needed to have everybody write the processes down, then I needed somebody else to literally write it and rewrite it for clarity and simplicity. Then I needed to have those people, the experts, review it. And then I need to have a beta group see if they understand it. What I'm doing now is I'm compressing that iterative process of getting to simple and clear. But all of the other steps are still really important.
So you're letting an expert come in at 301 level or whatever it is, and you're using AI to be like a translator to just simplify it down.
Yep.
That is so cool. Nice. I love this. That's such a good use. What tools are you using to do this?
I'm just curious chatgpt right now?
Yeah.
Yep. And it's amazing to do this and it. And it's really making this whole project, which we're just starting, we've been, we're like, we're just putting together the framework, we're taking a couple of our knowledge communities, our people and culture and our marketing to really dive in. First, we're empowering our community managers. So this is another thing. So for all of our knowledge communities, we have a community manager and they are responsible for ensuring that the knowledge that's shared within their community is current and relevant. And now we're adding simple and clear and we're empowering them with the tools to be doing some of this iterative process themselves. Because I would need like a dozen people in TAM if I was going to take on a project like this to move through a database as quickly as I feel we need to move it, at least to hit the high priority. So, for instance, my people and culture team now, they're going through all of their knowledge pages, all the wiki pages, and they're marking high priority. And then they're going to say, okay, we're going to work through four and it's going to take us two weeks. And then they're going to start creating new wiki pages right in synthesis, in the sandbox, and that's going to help again, speed the process up. So we're giving somebody the tools and the knowledge. And my community manager of people and culture said, him, he said, this is going to sound weird. In the intro meeting to this yesterday, he said, are you going to create a wiki page on how to create wiki pages? And I said, 100%. We've already got it in the works. Right. So that paragraph, what is the intention of that paragraph of the. Why other people can start trying to write it right now. They're giving me bullet points of it and I'm putting it together. I can empower them to do it for themselves. So then we'll have people embedded. It's embedded in their role, they're embedded in the teams. The teams are now empowered to own their own knowledge. And the teams are empowered to understand how their knowledge is being reformatted and represented to better impact our senior members.
So let's say Evan is in the design technology team and he creates some new tool that helps with shadow studies or something like that. And you want to get it added to Newton to your knowledge base. What's the new process? Do you interview Evan to get it out and then kind of dumb it down. Does Evan write a draft? I shouldn't say dumb it down, simplify it. Evan, you're so brilliant. My explanation was way too smart, Way too smart. Let's, let's was kind of firmed up for normal humans. Does Evan write first and then you guys review it? Do you interview and pull out of it? Like, how do you imagine that knowledge creation will work in this new paradigm?
So right now I'll give you two like perfect examples. So my software developers just created a fee calculation module. They developed it and so instead of our clunky Excel spreadsheet for us to create our fee proposals, they've literally created our own software to do it. So they wanted to beta test this out with project managers and project coordinators to, you know, get proof of concept and before we do a massive roll up to the whole studio. So they wanted to write a page to teach this beta group how to get in, how to create a project, how to add the time, like everything. And because that team hasn't been taught how to literally go into the sandbox and create wiki pages and they don't really have the same understanding of what we're trying to do. My KM team, somebody sat with them, interviewed them, got the intention, put together, the draft went back to them, they worked on it back and forth together. And then right now it was coming to me to do the sort of simplification process with my people and culture team because they've had more experience. And my community manager there, Jose, has had a lot of experience working in the sandbox, creating wiki pages, and he understands the intentionality because we worked through a few brand new guides. We did a role framework guide and we did a performance management guide. So they're a step ahead. So now they are actually doing the first step, which we've called discovery. So instead of having a discovery meeting with one of our KM specialists, they're doing their own discovery in their own department. They're going to create their draft guide, wiki page, whatever it is they're working on, and then they're going to meet with the KM team and review it at that point. So that's, you know, my, my end ultimate goal is that all of my teams will be able to do a lot of that discovery and draft layout work on their own. So then my KM team and my writing team and everybody else can really stay in their lane. And that's how we're going to tackle a project as big as this.
