When I look around, I sometimes get the feeling that, when it comes to AI, multiple realities exist at the same time. Some people still work mainly in the chat window, using ChatGPT to edit text, conduct research, or brainstorm solutions to a problem. Others connect multiple systems, build agents, automate workflows, or let task chains run largely on their own.
There’s more than just a technical difference at play here. I wonder if a whole series of other obstacles are coming together right now.
Perhaps some people simply don’t have any idea what they could actually do with these new possibilities in their own work. You can use ChatGPT quite effectively and still not know which task you should tackle differently next. The examples you see are only of limited help: If you don’t work in PR, a perfectly automated PR workflow won’t do you much good, and if you don’t develop software, the next coding agent won’t be of much use to you either.
Then there’s the technical hurdle. From the outside, you see the finished result: someone has connected several AI systems, linked data sources, and automated a process. What you don’t see are the decisions, services, interfaces, failed attempts, and possibly even the costs behind it all. This makes it hard to gauge whether something like this would be an afternoon’s work, a small project, or a new hobby for you.
And anyone who’s already using several AI tools may face a completely different problem. In that case, it’s no longer about learning yet another one, but about figuring out how to make it all work together effectively: Which system should I use for what? Which ones should work together? What’s worth automating, and where does the effort outweigh the benefits?
On top of that, there’s a question that’s probably even more fundamental: How does AI actually change my work if I don’t just view it as an additional tool? Maybe I’ll suddenly be able to do something myself that I used to have to outsource. Maybe my clients will be able to do something on their own that they previously needed me for. Perhaps a new service will emerge that couldn’t have made sense before. Or maybe surprisingly little will change.
Perhaps that is precisely what is so strange about the use of AI: The differences don’t just stem from who knows which tool or has greater technical skills. Some have a very specific reason to build something new. Others, while they see all the possibilities, haven’t yet found a convincing starting point for their own work. Still others have long been using multiple systems and first need to figure out how to turn that into a sensible workflow.
So what looks from the outside like “some are further along, others aren’t yet” could actually be something entirely different: very different starting points and very different questions.
And perhaps that also explains why conversations about AI so often seem to go at cross-purposes. Someone who’s just trying to figure out what might be relevant to their own work beyond the chat window is focused on something entirely different from someone who’s already thinking about agents, interfaces, and automation. Yet both say, “I work with AI.”
I’d be interested to know if others perceive this simultaneity in the same way. Feel free to email me or message me on LinkedIn about it.
Andere Use Cases helfen nur bedingt. In KIOA arbeitest du an deinem eigenen: Du klärst, was sich für dich zu verändern lohnt, probierst es zwei Wochen im Arbeitsalltag aus und entscheidest danach, wie du weitergehst.