How AI Could Truly Work with Us, Instead of Just for Us

There's a growing buzz around AI, especially as it becomes a staple in our everyday gadgets. But let's be honest, the real story is about how we make these tools truly work with us, not just for us. Practitioners are scratching their heads over how to design AI systems that don't just function, but actually foster trust and usability.

The Missing Middle

High-level principles like 'be transparent' and 'maintain user control' sound great at a conference, but on the ground, they're more like corporate mantras than practical advice. On the flip side, low-level UI patterns don't offer the strategic guidance teams need. What we're missing is something in the middle, a bridge if you'll, between these two extremes. Enter a new framework born from an analysis of 60 commercial AI apps. It's an attempt to carve out this middle ground.

A Framework for Collaboration

This new framework introduces the idea of human-AI coordination as three-dimensional: salience, involvement, and activity. Salience is all about how visible AI is. Involvement deals with what users can actually do with AI, and activity focuses on what AI does on its own. These dimensions aim to create a more cohesive interaction between humans and AI systems.

Real World Impact

The framework isn't just theoretical. It offers tangible tools like 'coordination zones' and input taxonomies, even charts that map out user journeys. It's a toolkit designed to help designers not just dream but build systems where AI feels like a partner rather than a tool. And let's face it, how many times have we seen management buy licenses without telling the team? This framework could help fix that disconnect.

Why Should You Care?

So, why should you care? If you're tired of AI systems that seem more like hurdles than helpers, this framework might be the fresh air you need. It's about time we moved beyond buzzwords and started focusing on real usability and trust. If AI is going to be a part of our daily lives, it should at least aim to make them better, right?

The gap between the keynote and the cubicle is enormous, but narrowing it might just start with frameworks like this. The tech world needs to get serious about human-AI collaboration, and maybe, just maybe, this is the first step in that direction.