The most interesting thing about my first driverless taxi ride in Shenzhen was how quickly it stopped feeling extraordinary.
I booked through an app, walked to a designated pick-up point and completed the verification. Inside, I fastened my seatbelt and started the journey from the passenger controls. There was no driver in front of me. Yet, after the initial curiosity, I was doing something very familiar: travelling from one place to another.
That experience left me thinking about the distance between a technology working and people making it part of their lives. The engineering achievement matters enormously. The service around it determines whether people can actually use it.
The customer wants to get somewhere
Developers understandably talk about what their technology can do. Customers usually begin with a less ambitious question: will this help me with something I already need to do?
A passenger wants a convenient journey. A small business wants to answer customers reliably. A manager wants to understand a problem before making a decision. None of them should need to become an expert in the machinery underneath.
Familiarity helps. In Shenzhen, booking the ride resembled using an ordinary ride-hailing app. I did not have to learn an entirely new behaviour before trying the new technology. The experience borrowed a habit I already had.
That is a useful lesson for an AI business. A powerful system may be easier to adopt when it appears inside the work people already do, with a clear task and understandable controls.
The small steps are part of the product
Between discovering a service and completing a task, there are many opportunities to lose a customer. An unclear instruction, a confusing handover or an uncertain next step can make an otherwise capable product feel difficult.
For a robotaxi, the journey begins before the vehicle moves. Can the passenger find the pick-up location? Do they understand how to enter and start the ride? Can they tell when it is safe to leave?
The equivalent questions arise in a workplace. Who supplies the information to an AI tool? Who checks its output? Where does that output go? If something looks wrong, whom does the employee ask?
These details deserve attention during an investment review. A demonstration shows what is possible under particular conditions. Repeated use shows how well the whole service fits into a customer's day.
Trust needs a way to handle exceptions
My ride was an individual experience, not a safety study or a test of the operator's economics. A service must earn confidence across many journeys and conditions, including the occasions when things do not go according to plan.
I would ask how a customer can get help, what happens when the system cannot complete a task and who takes responsibility for resolving the problem. A reassuring interface cannot substitute for a workable answer.
For business AI, this means setting sensible boundaries and making human review practical. People need to know where they can rely on the tool and where their own judgment remains necessary. Trust becomes easier to build when those boundaries are visible.
Convenience is different for different people
Removing a human role can also remove assistance that some users value. In my original account, I considered passengers carrying luggage and people who might need help getting into or out of a vehicle.
A service can work well for one group and still leave another poorly served. Before calling a product ready for everyone, it is worth asking who has been assumed to be the typical user and whose needs have been overlooked.
The same applies to workplace technology. An experienced employee, a new recruit and someone working in a second language may need different support. Good implementation takes those differences seriously.
Look for the work that makes adoption possible
For IJK, this is a useful way to assess technology opportunities. Alongside the underlying capability, I want to understand the customer problem, the cost of serving it and the work required to make adoption repeatable.
That work can include product design, integration, training and dependable support. It may attract less attention than a technical breakthrough, but it can make a substantial difference to whether the breakthrough creates lasting value.
Education belongs in that process too. Through UBI, IJK has a way to connect this discussion with learning: helping people ask better questions, practise with new tools and understand the decisions that remain theirs to make.
A new technology becomes part of daily life through many ordinary, well-designed encounters. My taxi ride made that idea tangible. The next question for an investor is whether a business can deliver those encounters reliably, for enough customers, at a cost that makes sense.
Adapted and expanded for IJK Capital Partners from Look, no hands! on Kenny Ng’s personal blog.