Why Click When I Can Say?

Why should I click when I can say? We use code and/or code use. With AI we make AI AI.

A click is a guess. You see a pixel on a screen that someone else drew there, and you point at it because they asked you to. You did not name the thing, you did not describe the thing, you just hit the target that was placed in front of you. The click is the user obeying the interface, not the interface obeying the user. Every click is a vote for the designer of the box.

Saying is the opposite. When I say what I want, I name it in my own words, and the work of translating the name into action is a work of the code. Like any other works you can become master of and develop memory around it. If the code can not take my words and produce my outcome, the code is incomplete, not my speech. I can not develop my speach without using it even if I accept having a poor speach. The burden belongs to someone, preferably a machine. I speak, the code follows, the code runs. That is use and dev in a single loop. Code uses my words to develop my words. Without a GUI in between pretending to be a protocol.

We use code and/or code use. Code that I can only click is a code that uses me so it also develops me! Code that I can say things to and get my outcome back is code that I use so I should be able to develop it too. The difference is who is driving. A button is a promise made by the developer to the user about what they are allowed to ask for. A sentence is the user telling the developer what is missing! One scales to whoever paid for the designer’s time. The other scales to whoever has something to say and something to gain from saying it. The first is a tax on the user, the second is a gift from the user. The first is a box, the second is a revolution.

AI does not change this principle. AI raises the stakes of it. With AI, saying becomes cheaper than clicking for the first time in computing history, and the gap between the two widens every month. If you still click when you can say, you are paying a tax in time, in precision, and in ownership, to somebody who drew a box around what you are allowed to want. That tax compounds, and the compounding is the whole point of the anti-code class. Keep them clicking and they stay renters of their own intent. This is an extractive AI.

With AI we make AI AI. Artificial intelligence becomes actual intelligence only when it is grounded in something that the user owns, on a box the user owns, under a license the user owns, paid for in money the user owns. Without those, AI is a larger click, a more expensive guess, a prettier GUI for somebody else’s business. With those, AI becomes the translator between what I say and what the code does, and that is the translator we have been waiting for since the first terminal. The A in front becomes silent when the loop closes: I say, the code runs, the outcome is mine, the next sentence extends the last. And I can share these. That is how we learn words from the one who has something to say not only from the one who has something to extract.

So the test is simple. If you must click to get your outcome, the system uses you. That’s. If you can say it and the system produces the outcome under your ownership, you use the system. Everything else is decoration around this. The clicks that remain are clicks on code you wrote yourself, and even those should fall away as soon as you can say the same thing faster.

Set a time that you don’t click on the screen after that which essentially means you need to make your own screen. Life becomes easier for you and harder for the people who want to extract from you. Win to lose them.