Episode V of Phase One: Tech comm predictions for 2026
Fifth episode of the AI & Docs podcast series is up! In this one, Tom and I discuss our predictions for tech comm in 2026, the evolution of writers into automation engineers, the risk of the Reverse Centaur dynamic, and the growing value of authentic human connection.
You can watch / listen to the episode here:
Some of the things I said:
On regulation and Phase One
"We are still in Phase One. So, you know, and in this phase is like when any new tech is introduced there's a period of time where legislators are kind of clueless. Things haven't happened yet or are still happening and decisions haven't been taken yet. And law usually moves a lot slower than tech. But then it catches up. And I think it will catch up eventually."
"And I'm not talking just about regulation but also I think I hinted at that in the post is pedigree. We will want to know where the models come from and you know, who are their daddies, you know, father and mother, who created them, what data they used to train the model. Because that will… we will have like model connoisseurs, you know, much like it happens with wine."
On the reverse centaur
"The Centaur is like the best scenario. It's like human upper torso with the body of a horse. So you go fast but you retain your human intelligence and control. The reverse Centaur is like well essentially you have the worst of both worlds. So you are like an approver, like just a machine… like a human appendage of AI. AI is doing the stuff and the human is just clicking a button saying 'Yeah go ahead.' And it's really demeaning of what is human."
"My take is still optimistic in that we can avoid that mostly for the most part. Or at least sometimes we'll be reverse Centaur and it's fine. Some other times we'll take more control. You know, it's something like that."
On systems thinking and automation
"I think automation engineering is like an expression… one of the many expressions of systems thinking. Which I believe we brought up in previous podcasts as well. And I think systems thinking is one of the top skills I would probably seek in anybody in the future working with AI. Because it's really about orchestrating and seeing the big picture and relating systems between each other. Which is something that I don't expect AI ever to be able to do."
"And wanting to do something is purely organic and human. Like you want to achieve something that you care about."
On the human element
"I think technical writers will be like therapists of AI. Or you know like the non-player characters in massive role playing games, you know like the shopkeeper. Because I expect people at some point will want that truly human contact. You know of someone that explains tech to them."
"And I think Apple has always had these human touch in their design system. And I think they're still trying to figure out how to really make it like a very human experience like even an emotional experience. You know if they succeed we will probably… you remember the movie Her from 2013? I mean that's like how I picture like Apple AI going forward. But I don't know. Like that's a dream. That's a dream you know."
On fighting back
"I was like well you know let's you know let's go on the offensive. Let's let's write something to that people. Let's address those folks who are taking those decisions."
"And ironically I was using Manus AI to to like try some layouts and it you know when I showed the post without saying anything Manus created this single page application that that then I used to in /reconsider. And I thought well this is brilliant… But isn't it ironic that AI helped me design that you know like it helped me get the message through that yeah of course you don't want to use me like let me help you with that."
"I'm seeing more and more technical writers blogging these days and it's really encouraging. And I think this is way better than SEO. I mean we are planting the seeds in the future LLM versions of you know what they should be doing and and and telling stakeholders about. So yeah get get your words out there get your voice out there because that's the context that is gonna you know upon which they're gonna train the models in the future you know."