bernardlunn

First AI came for journalists, then investors then programmers

If you run an AI business you want to replace expensive humans who are not protected by regulators.

Journalists are both expensive and not protected by regulators. So this human was not surprised to have LinkedIn suggest I upgrade to premium so I can have AI rewrite my post. I am retired so not interested, but imagine a big media business making this decision.

Investors are better protected by regulators, so grab your popcorn for this one.

Many programmers are in denial about AI replacing them. Visions of making a fortune writing AI are part of this denial. First programmers will be commodified by AI, as the few humans still making decisions will not be motivated to learn how to instruct AI.

First they came for the socialists, and I did not speak out— Because I was not a socialist. Then they came for the trade unionists, and I did not speak out— Because I was not a trade unionist. Then they came for the Jews, and I did not speak out— Because I was not a Jew. Then they came for me—and there was no one left to speak for me.

I am very pleased that trust busting is back in fashion

This Baby Boomer was much influenced by Small Is Beautiful: A Study of Economics As If People Mattered, a collection of essays published in 1973 by German-born British economist E. F. Schumacher.

Peter Thiel, who is vastly wealthy, wrote in an otherwise great book (Zero to One) that “competition is for losers”. If you are a tech entrepreneur you probably like competitive markets, as they enable your company to thrive.

Digital economics is different. In the analog age, competition was measured by prices and as long as there was some competition, economies of scale meant a dominant firm could offer good prices.

Digital economics is different in two main ways:

1. The dominant player usually has market share of 90% or more which is really a monopoly - just ask anyone who tried competing. This happens due the power law aka the more people who use a service the more valuable it is.

2. The digital monopoly often makes money by owning your data and attention. The analog firm has to advertise to get your attention.

These two differences explain why trust busting has been in and out of fashion. During most of the time when trust busting was out of fashion, there were few digital monopolies but today is different and as a tech entrepreneur I like competitive markets. That is why I am very pleased that trust busting is back in fashion.

The mistakes would be noticed by any human paying attention. For one hilarious example, a video about WW2 mentions Herman Goring but the transcript shows Herman Gary.

No content creator would make such a rookie mistake.

Maybe this is offshore humans. Anyone know?

Let’s use a simple job like transcribing audio files as an example. 100% by AI leads to simple dumb mistakes that any human can see. Today human users tolerate such errors, but will probably switch to a better service as soon as it is available.

The economic game changer is obvious - a massive cost cut. This is not quite 99% as the AI will be used by a human who has to be paid and the AI usually has some cost.

In short it is AI and humans not AI or humans.

The geopolitical game changer is not so obvious. A simple job like transcribing audio files has typically been done in offshore locations. The future is likely to be AI and humans, but the humans are likely to be highly skilled and expensive and there won’t be many of them. Some of these skilled/expensive people will be in offshore locations, but that is likely to be a rarity. Most of these jobs will be reshored. This will be a problem for countries who offer lots of low-skilled/inexpensive people via offshoring. These countries will be at a geopolitical disadvantage in an AI led future.

What has been your experience with this?

No, that is not part of a pitch to VC. It is my name for content services that have ads for services including a premium version that you pay for directly and which do not have any of those “annoying” ads (of course they do not use the word annoying).

The Internet needs a sustainable ad model that users/consumers also find to be useful not annoying.

Financial results will be positively impacted in the short term by Annoyance As A Service, but this leads to a spiral of doom. The annoying ads become more frequent, leading to a short term financial boost but also to users/consumers leaving the service which leads to the annoying ads become more frequent which leads to…

Have you noticed this in any services you use or work on?

It is what separates great entrepreneurs from average ones.

People often do not do something even though they believe it is the right thing to do, simply because they believe change is impossible.

The wonderful thing which great entrepreneurs know in their bones is that you only need to convince a few people that change is possible. Those few people will convince others who will… anybody with a basic sense of math will see how this translates to mass acceptance and market share.

The greatest entrepreneur ever - Steve Jobs - had total mastery of this.

Amara’s Law (coined by Roy Amara, scientist, futurist and President of the Institute of the Future) is:

“We tend to overestimate the effect of a technology in the short run and underestimate the effect in the long run.”

This is true for game changers such as Internet, Blockchain & AI. Some hyped technologies never make it through all five phases of the Gartner hype cycle:

1 Technology Trigger

2 Peak of Inflated Expectations

3 Trough of Disillusionment

4 Slope of Enlightenment

5 Plateau of Productivity

So it is possible that Blockchain & AI never make it big in the real world. The Internet has clearly made through all five phases. However I believe that Amara’s Law will also be true for both Blockchain & AI. While Blockchain is clearly deep in the Trough of Disillusionment, AI might be just starting that journey after spending 2023 at the Peak of Inflated Expectations.

What do you think?

Personally I don’t use either platforms although I used to be a bit addicted to Twitter until Musk destroyed it imho.

There are only two online content monetization models - ads/indirect or direct and both have pluses and minuses.

The news about Substack giving Nazis a platform shows us that direct monetization is not the answer, maybe until a really simple low cost micropropayment platform emerges, so we can pay for articles we like. Until somebody manages to offer privacy friendly personalization, the ads/indirect model will not be the answer.

What do you think?