Charles Simonyi compared AI to an unusually generous goldfish: it may grant many wishes, yet eventually the user could run out of ideas. The central question is therefore not only what the machine can do, but what a person wants to do with it.
Machine speed makes human intention more valuable
Generative AI can draft text, explain concepts, write code and organise data within seconds. Fast output does not replace the work of deciding who a solution is for, which problem it should solve, and what would make it genuinely useful.
The same tool can open very different paths. A family photograph may lead to historical research; an odd data point may suggest a new investigation; a repetitive task may inspire a small custom application. For someone else, the same service may remain another form of entertainment after a few amusing images or conversations.
An old knowledge gap in a new medium
Greater access to information has never automatically erased differences between people. Books, television and the internet all expanded opportunities for learning and entertainment at the same time. Those with prior knowledge, sustained interest and experience interpreting information were usually able to turn the new supply into greater advantage.
AI enters this longer history. Access matters, but it is not the same as benefit. Attention, persistence, source checking and the ability to connect new information to existing experience still shape what a person can do with the technology.
Creation or easy entertainment
A chatbot may take over the evening hour once occupied by a television series, a mobile game or short videos. That does not make recreational use inherently harmful. Its effect depends on what AI replaces, how much time it occupies, and how it changes work, learning or human relationships.
The divide is not between two fixed groups. The same person may use AI for work one day and for play the next. What matters is the balance and the next step: does the finished answer lead to another question, a check, an experiment or something genuinely made by the user?
Prior knowledge remains an advantage
A polished AI response can easily look like completed work. Yet one confused name, missing source or false inference can send an entire project in the wrong direction. Users must recognise what needs verification and when another source or a human expert is required.
Prior knowledge does not make AI an exclusive tool. It makes the tool easier to direct. Someone who understands a field can ask more precise questions, notice uncertainty, compare answers and integrate the output into a broader piece of work.
AI can increase the return on curiosity
AI offers particular value to people who already want to make something. A teacher can build exercises for a specific class, a small business can analyse order data, and a local community can turn scattered documents into a searchable collection. Parts of the work that once required specialist help or a separate budget can become faster and more accessible.
Curiosity here is practical behaviour: tracing a claim to its source, reading the referenced study, testing a suggestion in a real setting and returning with a better question. Used this way, AI does not end the thinking process; it helps move it forward.
What might the twenty-sixth wish be?
The twenty-sixth wish may not come from the machine. It can emerge from an interest that existed before AI was switched on. After processing family letters, someone may want to understand the history of the town in which they were written. After building a catalogue, they may see how other collections could be connected to it.
AI can democratise specific capabilities, but it does not automatically equalise motivation, knowledge or patterns of use. Institutions therefore need to do more than provide access. They must teach problem framing, source criticism, verification and constructive use.
The technology may be available to everyone. Whether it becomes entertainment, quick assistance or the starting point for a new study, product or enterprise still depends on human choices.