I had a really interesting chat earlier this week with a founder from a YCombinator startup. We talked about various industry topics and AI in particular. Considering my background, he was curious to hear my take on AI tools and how they’re transforming businesses, especially when it comes to UX and Product Design. Since it was such an interesting topic and very relevant at this time, I thought I’d share my thoughts here too!

TL;DR – It’s a whole lot of hype and little substance.

The longer version –

At its core, UX is about de-risking your product or service by understanding your user and how they use your product or service, and designing your solution accordingly, so it solves their pain points or addresses opportunities for them in their natural context of use alongside the real-world constraints that they face.

Right now, the current generation of AI tools can create some pretty slick interfaces and cool prototypes with just a few prompts. But here’s the thing: they’re completely reliant on the data that they have been trained on, which might not even be relevant to your target customer segments. It’s a bit like when company leadership wants to design a product or service, saying that they know their customers and there’s no need to do user research, but as we’ve all seen, that is rarely, if ever, true, and there is a disconnect between what they know and customer pain points. So, products and services created without user research rarely result in great customer experiences and often require expensive rework.

AI tools in this generation (and for at least the near future, in my view) are quite far away from being able to actively observe and analyze user behavior, conduct deep dive interviews to understand pain points, and process that into meaningful insights that can then be used to build the right solutions for your target customers.

That said, AI-powered design tools are absolutely awesome for productivity gains in the design process. I would call them the new whiteboard tools. No longer do you need to spend a bunch of time on whiteboards, Figma, and prototyping tools to figure out if concepts are feasible to build out. With a few simple prompts, now we can rapidly prototype concepts alongside engineering and business teams to determine feasible concepts that we can then proceed to test out with customers.

Now some folks might say that “We could just launch and validate it in the market. We don’t need to do any research prior to launch”, and yes, you could do that, but that misses the entire point of de-risking the product. You risk significant brand and reputation damage if customers don’t like your product. You must capture feedback, fix issues, and relaunch. If there’s a major change, you’ll need to help customers navigate it, which consumes more resources. These add more cost (time, money, resources) and lost sales revenue during this time. So why inflict this pain on yourself and your company unnecessarily?