In 1996, Garry Kasparov lost to a computer for the first time. A single game in a match he ultimately won, but the moment registered differently than a normal loss.
The following year, he lost the full match to Deep Blue.
What he did next is the most important part. Kasparov created Advanced Chess, a new format where humans played alongside AI as collaborative partners. It spread. A decade later, freestyle chess tournaments produced a striking result. The winning teams weren't the strongest grandmasters, nor the most powerful computers. They were amateur players who were best at collaborating with their machines.
Weak human plus machine plus better process beat strong human plus machine plus inferior process.
Product design is living through the same moment right now.
Research from Anthropic on long-term AI usage surfaced a pattern worth sitting with. People who had spent at least six months working with Claude dramatically outperformed newcomers on the same tasks. The gap had nothing to do with intelligence or better prompts. It came from something harder to measure, an intuition for when to lean on the AI and when to override it.
That intuition only builds through use, and use here means six months of being mediocre with a tool before the compounding starts.
TechCrunch has called this the "AI skills gap." That undersells what's happening. A skills gap implies something you can close with a workshop and a lunch-and-learn. What's forming is harder. From a distance the gap looks like curiosity, since the people pulling ahead are the ones who try things. Up close, most of what's compounding is tenacity, the willingness to stay with a tool through the awkward phase long enough for intuition to form.
Three skills are doing the heavy lifting in this era. Curiosity, tenacity, and adaptability. They get discussed as if they're interchangeable. They aren't.
Curiosity is what you choose to try. Adaptability is whether you can change shape when the tools do. Tenacity is what keeps you in the work during the months you're bad at something new.
Tenacity carries the most weight and gets underrated most consistently. Curiosity gets you to start. Adaptability protects you from getting whipsawed by change. But the people who actually pull ahead are the ones who can be visibly mediocre with a new tool for long enough that mastery becomes possible. Most people quit two weeks in. Tenacity is the muscle that decides who's still there at month six.
Carl Rivera, Nubank's new Chief Product Officer, said his goal for 2026 is "to be bad at things." He meant it as genuine strategy. The only way to develop intuition with new tools is to use them before you're good at them, tolerating the discomfort of not knowing what you're doing long enough for the intuition to form.
There's a concept in learning science called "desirable difficulty." Some amount of struggle improves retention and understanding. Things that feel easy to learn are often easy to forget. Things that require you to figure out the path yourself tend to stick. AI tools are getting easier to use every month, and that's mostly a good thing. The default path, the "good enough" way to use them, is also getting more comfortable. Comfortable corrodes tenacity faster than almost anything.
Carol Dweck's growth mindset research demonstrated the same dynamic in educational settings. People who believe their abilities grow through effort approach challenge differently. They persist longer, interpret obstacles as information, and practice in harder, less comfortable ways. The orientation itself produces the durability.
We've been running workflow reinvention sprints at Nubank to rethink how we work as an AI-first company. Three-day focused blocks where a team does nothing but rethink, and redesign, how their work happens. The pattern is consistent. The people who get the most out of them aren't the most senior. They're the ones who showed up willing to look stupid for three days. The sprint gives them a container. The willingness was already there.
Tenacity, properly understood, is the capacity to stay in a state of not-knowing long enough for the knowing to form.
Curiosity is what loads the question. When a new tool shows up, most designers ask, "What does this do?" They watch the demo, learn the prescribed workflow, and become competent. They use the tool the way they were shown, in the contexts they were shown, and not one inch further.
A smaller group asks a different question. "What could this do that nobody's tried yet?" They poke at the edges. They use the tool wrong on purpose to see what breaks. They stay up too late on a Tuesday because they had an idea and wanted to test it.
I've watched this play out across industry. The designers pulling ahead aren't the ones who were already the most technical. They're the ones who were already the most curious. Before AI, they were the first to try a new Figma plugin, the first to learn a new prototyping tool, the first to wander into an engineering standup just to see how the other half worked.
Curiosity without tenacity is dabbling. The Tuesday-night experimenters who matter are the ones who returned on Wednesday, and Thursday, and the Tuesday after that, even when nothing useful had emerged yet. AI made the gap visible and started accelerating it. The gap itself predates the tools, and what's compounding behind it is the willingness to stay long after the willingness to try has faded.
Adaptability is what keeps both of those skills useful when the tools change.
John Culkin, writing about Marshall McLuhan's ideas in 1967, said it plainly. "We shape our tools and thereafter our tools shape us." From stone to iron took thousands of years. From desktop computing to mobile took roughly a decade. From whatever is current in AI to whatever comes next may take weeks. The tools are changing faster than our ability to settle into any single way of working with them, and that acceleration is itself accelerating.
The only thing we can say with certainty about AI tools is that they'll change. We can't predict which capabilities emerge next or which interfaces designers rely on in 2028. Adaptability lets the muscle you built on this generation of tools transfer to the next one.
Adaptability without tenacity is reactivity. People who change shape constantly without ever staying long enough to develop intuition end up performing fluency without ever building it. They learn the surface of every new tool and the depth of none. The designers who shape what "designer" means in three years will be the ones who can do both, change shape when the tools demand it and stay put when the work demands it.
Kasparov wrote about Advanced Chess years later. He said the key insight was what he called "the process of collaboration." Knowing when to defer to the machine and when to override it. When to trust the data and when to trust your gut.
The best players were the ones willing to be wrong, willing to be surprised, and willing to keep playing badly for long enough that they got good.
That's what the gap is measuring. Curiosity gets you to start. Adaptability keeps you upright when the tools shift. Tenacity, more than either, decides who's still here in a year. The designers who build a durable edge in the AI era will be the ones who decided, without being asked, that staying with the discomfort was their job.