We dont get AI

In a sense, we are back in the 1890s in USA.

Phonographs with coin slots are being set up on streets and in parlors. They are hugely popular, people flock to them. You insert a coin, and the U.S. Marine Band starts playing when you press a button. Yes, it is you that start the band. Even more startling, the band is not there, and in many cases there wouldnt have been room for it.

The sound quality is horrible, but the experience is amazing nonetheless. Some other popular recordings are artistic whistling and jokes.

Up until now, talking has been a defining feature of a human, and then come these machines that talk. To people, that must have felt like walking down a staircase that was missing a step. Previously sound and the source producing it, were always simultaneous. But now audio recording disconnects sound from source. This is such a monumental new thing.

People want to know about it, and they want to experience it. Some are disappointed in the sound quality. Some are skeptics, but they are the exception. There is a general sense of optimism.

Take note.

People talk about phonographs, papers write about them, demonstrations showcase them. Everybody has heard about them.

There are stories about people hearing a recording for the first time, and thinking there is an orchestra next door. But there is no evidence that ever happened. People understand more than they are later given credit for. It is really not that hard, it is only a recorded sound that is played back.

Take note.

Phonographs are for the rich. If you want one, expect to pay what amounts to your monthly salary. They are promoted through ads with poetic slogans like "talking machines", "voice photography", "mirror of sound" and "the instrument that sings".

Edison himself, back when he started to work on the phonograph, called it a "speaking telegraph". His ideas for what it could be used for, were generic and a little bland. He might just have struggled to put it into words.

Take note.

New business areas have started up. Production of phonographs, production of phonograms (the records) and selling them. But there is more.

Some guys rent a phonograph, and sell tickets to their "Phonograph Concert". People pay to come and hear them play phonograms. That does make sense, even though it sounds scammy in todays ears.

Others make "voice portraits" of prominent persons. Others experiment with using audio recordings to induce hypnosis. Others again will do a voice diagnosis based on an audio recording of a persons voice, resulting in a profile for temperament, illness, criminality and more.

Take note.

Pattern

Did you notice the pattern?

People are genuinely curious about the new thing, and want to experience it. It makes sense for them to learn more about it. Maybe they see the potential it has, or maybe they just feel it subliminally. There is a general, maybe slightly nervous, sense of optimism.

People in general seem to have a basic understanding of the new thing. They understand what it does, and what it is. This basic understanding quickly reaches a level that is comparable to what it will be later, long after the new thing has been adopted.

The ramifications of the new thing are not so well understood. It is seen in the light of well-established technologies. It is seen through analogies. It is difficult for people, including specialists and nerds, to see how it is going to change their lives.

There is a flurry of trying out all manner of ideas for how to use the new thing. Some of it is magical thinking and crazy stuff. The general sense of optimism provides some of the driving force. This phase starts strong and take decades to settle down. The ramifications of the new thing gradually appear as more and more of the ideas are successful.

This pattern apply to adoption of any major new technology. It applies, as we saw, to audio recordings. But also to aeroplanes, to plastics, to nuclear power and so on.

AI

The pattern also applies to AI.

You are in the middle of it, and are probably engaged in it in some way. Try to put that aside for a moment. Take a step back and look at the whole thing as an observer. You are observing an interesting phenomenon first hand.

Bullet one. People are interested and a little nervous. They are seeing the potential of AI. Curiously the major problem people see is that AI can take their job, which still is to see the potential of AI, but used against themselves (while everything else is held constant).

Bullet two. People in general have a basic understanding of what AI does, and somewhat of what it is.

Not how AI builds up its weights, and how they are used to generate output. That whole thing is quite involved. Only a small number of people understand AI on that level.

Dont expect the basic understanding to change significantly. It will likely still be on the level it is on now in twenty years, or in hundred years.

Bullet three. It is unclear to everyone how AI will change our lives. People tend to see small bits of the big picture, for example that they might loose their job because AI can do it, or that their car can drive them around while they take a nap. Extrapolations of the well-known.

One thing I find remarkable is that when I ask young people about the future with AI, what I hear is that they are just as much in the dark as everyone else.

We dont get AI.

Bullet four. There is a flood of ideas and projects that involve AI in some manner. Lots of them seem a little opportunistic, to not say scammy. There are also crazy and semireligious stuff as well. We are clearly in the start of this phase.

Expect most of all this to fail. It will mostly be forgotten in the future, whatever that is going to look like.