TL;DR: The AI debate isn't settled, but your vendors aren't waiting for it to end. They're shipping AI features into platforms you're already using, whether the technology is proven or not. What lands in your environment is someone else's bet. And the users inheriting those features are either ignoring them or trusting them too much. Neither is great.
If you're seeing this play out in your own tools, please share this with someone who needs to hear it. The more people go in with eyes open, the better. Thank you :)
I’ve been watching the AI debate play out for a while now, and it hasn’t really settled. Every major release comes with big claims, then the pushback follows, and then the whole thing starts again with the next release. We end up in these loops; Hype versus reality, promises versus outcomes, these two forces, equally loud, going around in endless loops.
Which is fine. But I can’t stop thinking about what this means for the rest of us, while the arguments continue, the vendors building the platforms we run our businesses on, have all mostly, already made their call. They’re not waiting.
You can totally understand why. You can’t sit on the side-lines in a competitive market, hoping you made the right call by not doing anything. You don’t want to be the next Kodak. Especially while competitors ship AI features. So they do, like everyone else. FOMO is real. They integrate, they add features, update pricing, introduce credit and token models, and push things through, often before the technology has really been proven in enterprise environments.
Companies like Apple have been criticised, rightly or wrongly, but they also have the ability to hold back. They’ve done it before, waiting until something is genuinely ready before shipping it. Most SaaS vendors aren’t in that position, and honestly the market isn’t rewarding patience right now. It’s innovation at all costs.
So what lands at the enterprise level is someone else’s bet. We start to get feature creep. May even have pricing adjustments that shift to cover AI costs mid-cycle. New capabilities showing up in your environment whether you asked for them or not. Yet the debate about whether AI is ready is still happening. Whether these gambles pay off is anyone’s guess at the moment.
Then there’s the user side, which I find just as interesting.
When AI features arrive inside tools people already use every day; built into the workflow; the insights engine, etc, there’s an adoption question that plays out. Some people won’t touch these features. They’re cautious, or sceptical, or just comfortable with how they already work. It adds little value for them. From a vendor’s perspective that’s its own headache: you’ve built something a chunk of your customers are ignoring. Yet now it’s costing you and you must maintain it.
But the more concerning pattern is the flip side. Users who do adopt, and do so quickly, without much critical review. The output feels correct, after-all, we’ve been told this is PHD level capability. It’s fast, it’s confident, it’s hyped and it lives inside a tool they already trust. So they use it, move on, and review the outputs less than they probably should.
We’ve already seen this play out where consulting firms had submitted AI-generated work that was wrong. Not because people were careless, but because the tool and the hype was convincing and the habit of checking their work wasn’t in place. That’s what happens when technology moves faster than the judgement around it, and people quietly start trusting the output more than they should.
Where am I going with this? I’m not quite sure. I don’t think there’s a neat answer here, and I’m not going to pretend there is. Vendors will keep shipping because they must. The debate will keep cycling because the technology keeps changing. And in the middle of all that, enterprise operators are making decisions about platforms, pricing, and people, with incomplete information, whilst under pressure to be innovative or efficient, and to deliver tools that actually get the work done.
The hype debate is worth following. But so is what’s already sitting in your environment while the debate plays out. I guess this is a cautionary tale to go in with eyes open, understand where the features add value and where the risks lie.
Curious whether others are feeling this, the gap between watching the broader AI conversation and managing what’s already embedded in the tools you run every day. Thoughts?
~ Pete G


