Explore

Jason Barnes had wanted to be a professional drummer since he was a teenager. In 2012 an electrical accident took his right arm below the elbow. After a long stay in hospital and several operations, he set his mind on getting back to the drums.

He did not wait for someone to solve it for him. He built his own prosthesis, a spring-loaded drumstick, and carried on studying at a music school in Atlanta. It worked up to a point. He could play by moving his elbow, but without a wrist or fingers he could not control how fast the stick moved or how it bounced back off the drum. For a drummer, that bounce is everything.

His teacher introduced him to Gil Weinberg, a professor and the founding director of the Georgia Tech Center for Music Technology, who had already spent years building robots that play music. Weinberg's team made Barnes a robotic prosthesis with motors that power not one drumstick but two.

The first stick follows Barnes. Sensors read the signals from the muscles in his arm, so when he flexes, a computer tightens or loosens the stick and changes how it rebounds. In other words, the machine learns to listen to the person.

The second stick is where the story turns. It has its own small computer, which listens to the music being played, picks up the beat, tempo and how busy the playing is, and improvises a part that fits. Barnes does not control it directly. If he wants to play alone, he pulls it away from the drum. If he is curious, he lets it play and responds to what it does. Weinberg described it as having a mind of its own.

Notice what the team chose not to build. They did not build a robot to replace a drummer. In an interview with Vice, Weinberg said the aim was to learn more about how people create and improvise, not to take musicians' jobs. The machine does three things that are worth borrowing. It reads human intent from messy signals. It adds a partner that contributes something the person cannot. And it leaves the human with the final say.

The story also has an honest footnote about the distance between a lab and everyday life. MathWorks reported that the system was complex enough that each performance needed an elaborate set-up from the Georgia Tech team, and Weinberg hoped to redesign it so that Barnes could perform anywhere. The team has also worked on a shoulder-mounted extra arm controlled by brain signals, according to ASME. Good ideas often start out fragile and become practical in steps.

So here is the question to take away. Where, in your own solution, could AI listen to people and add a capability they do not have, while they stay in charge? The prompt below helps you find out.

Two years ago, Rin opened the RoboChef Restaurant with a fully automated kitchen. A ceiling-mounted system of robotic arms could chop, season, plate and cook with a precision no human kitchen could match. For the first few weeks, curious diners filled every table.

Then the novelty faded. Within months the dining room sat half empty most nights. The food was consistent, but in Rin's own words it was "correct and forgettable." Ordinary restaurants with human chefs were beating a restaurant built around cutting-edge robotics. One slow evening, Rin sat alone in the empty dining room, running the numbers again and wondering whether the whole idea had been a mistake.

A conversation in a nearby café changed everything. A friend asked a simple question: your system can carry out any recipe with perfect precision, so why is it only cooking yours? What if it cooked like the most celebrated chefs in the world? Rin realised the robots had been treated as a replacement for an ordinary kitchen, when they could be a vessel for culinary intelligence that no single kitchen could hold at once.

Rin brought in an engineer, and together they built a database of techniques, flavour signatures and plating philosophies drawn from renowned chefs across many culinary traditions. The aim was never to imitate one chef. It was to give the system an elite range to draw on, dish by dish.

Word spread. Within months the restaurant was booked most nights, and it was known not for its robots but for the breadth and precision of what came out of the kitchen.

Notice what changed and what did not. The robots stayed exactly the same. What changed was what Rin put into them. Call it unique intelligence integration: embedding specialised, elite-level expertise, or a genuine edge of your own, into a system or process instead of settling for its default.

The same is true far beyond restaurants. Any organization can buy the same tools as its competitors, and a tool everyone owns produces what everyone else gets. The advantage comes from what you teach it that others cannot.

So here is the question to take away. In your own solution, what is the default you are settling for, and what specialised expertise, data or capability could you feed into it? The prompt below helps you find out.

Three years ago, the Sustained Farm was a modest, manually operated growing unit on the edge of the city. A team of five managed planting, watering, harvesting and selling across a seasonal calendar that left them dependent on the weather, exposed to crop failure and worn out. Revenue was inconsistent, and two difficult seasons had pushed the business to the edge of viability.

After a hard week, the founder, Priya, asked her team a question. It was not "how do we grow more?" It was: what if we could produce ten times what we produce now, with half the people, in any weather, all year round? Her team said it was impossible. Priya said: let us find out why we think that.

That question is the heart of moonshot thinking. A moonshot is a bold leap, not a small improvement. Priya's team worked through it in four steps. First, they reframed the challenge as a ten-times improvement instead of an incremental one. Second, they identified the assumptions that made the current approach feel like the only option, and deliberately discarded them. Third, they identified the technologies that could make the ten-times improvement possible. Fourth, they redesigned the whole operation around those technologies instead of adapting the old model.

The assumptions they dropped were that soil was necessary, that sunlight was the only source of energy, and that human hands were needed for monitoring and adjustment. In their place they chose computer-vision crop monitoring, automated nutrient delivery and predictive AI yield optimisation, and rebuilt the farm around them.

In the story, the farm now operates vertically in a climate-controlled space with a team of two, and revenue has grown sixfold. It sells directly to consumers and corporate clients, and has become one of the city's most visited showcase businesses.

Most improvement plans start from the current method and ask how to make it a little better. Asking for ten times better forces a different kind of thinking, because no tweak will get you there. You have to question the assumptions the old method rests on.

So here is the question to take away. Which assumption does your organization hold about how your problem gets solved, and what would you build if that assumption were false? The prompt below helps you find out.