Hammers in Search of Nails

Alright, let's drop the pretense. That whole 'problem-first' purity we preach in UX? It’s a nice story, but it’s rarely the whole truth, is it? We’ve always been influenced by the tools at hand, subtly steering our problem-finding towards what our tech can actually do. Now, don't get me wrong, blindly chasing solutions – the classic 'hammer looking for a nail' – that’s a fast track to irrelevance. The sweet spot has always been that pragmatic middle ground: finding real user needs that our current capabilities can realistically address with high impact.   

But what happens when the 'hammer' that lands in our toolkit isn't just another incremental upgrade? What if it’s something so potent, so paradigm-shifting, that it forces us to re-evaluate everything, including that comfortable middle ground? What if this new hammer can smash through problems we once dismissed as impossible, problems that weren't even on our radar because the solutions were simply out of reach?

Large Language Models (LLMs) aren't just another incremental upgrade – they represent a trifecta of power, a combination of capabilities we've only glimpsed in isolation until now:

• Natural Language Understanding: They genuinely get what we mean, nuances and all.   

• Reasoning: They can connect dots, break down complex problems, and make inferences.   

• Tool Use: They can retrieve information from data sources, call functions, and access other applications.

When a tool unlocks this many capabilities, it's time to flip the script. Think back to the introduction of the airplane, the automobile, and the internet – these were technologies built largely on experimentation and exploration. The uses we found for these new tools were often things we couldn't have even imagined before they became part of our world. We had to discover how these new tools could revolutionize tasks and create new value.   

But how will we know when and where this new hammer can provide value? By swinging it. It’s in the trying, the tinkering, the occasional (or frequent!) surprising outcome that we truly learn its strengths and its current quirks – like its potential for hallucinations or biases if not carefully managed  – that we’ll discover the "low-hanging fruit" that was previously out of reach.   

Because here's a truth that often gets lost in the excitement of this new technology: our core mission as product builders hasn't changed. We still need to deeply understand our users, their context, and their problems, so that we can help identify the areas where this new technology can provide real, tangible value. It doesn't matter whether we call it an "Chat Bot" or label it "GenAI" or name it an "Agent". The only questions that matter is whether or not it solves a real user problem and makes sense for the business. As UX professionals, we need to actively hunt for the "nails" that our LLM hammers are uniquely suited for – problems we previously deemed unsolvable, or perhaps didn't even recognize as problems because no solution seemed possible.   

So, what does this mean for us in UX? It means our role is expanding. We need to balance our cherished problem-driven design approach with a healthy dose of technology-driven opportunity exploration. We need to get out there and explore this wild new technological frontier.   

This is our call to action. Let’s embrace the experimentation. Let’s lead the charge in discovering what’s now possible. The future of human-computer interaction is being written as we speak , and with LLMs, we have a powerful new pen – or rather, a very impressive hammer. Let’s get out there and discover all the amazing things we can build.