Now that you have some idea of what the book is all about, let us start our criticism of the book. Starting with the prelude and working our way through chapter by chapter. We will only deal with the philosophically interesting parts of the book and leave the others alone.
But why pick on Tegmark at all? Well, because a lot of other AI researchers agree with many of the ideas presented in this book.
This episode will deal with the prelude and will then cover chapter one and chapter two.
The prelude is a fantastical scenario where a bunch of programmers called the Omega Team covertly creates some pretty creates some artificial intelligence, which they call Prometheus. They then proceed to secretly unleash it upon the world.
Prometheus eventually proves to be up to any task given to it. It can make money on the stock market, make movies better than most human studios, it can even design and create amazing new technology. You name it, Prometheus can probably do it better than any human could hope to.
It massively disrupts the economy and starts dominating virtually every industry in existence. It and the army of AI’s it creates put countless people out of work and people all over the world start depending on Prometheus for virtually everything.
This allows it to expand massively, virtually without limit. Until it is entrusted with almost everything. Including political power. At this point, it essentially runs the world while humanity sits back and reaps the benefits.
While this makes somewhat interesting science fiction, I hardly consider it very realistic. But it seems that this chapter represents one of the possible courses of development Tegmark considers AI could take. And one of the better ones, at that.
However, this is apparently the kind of future Tegmark believers might happen. He is far from alone in this and he is joined by many, such as Larry Page from Google and others.
I am not so sure I agree with this. Nor do I agree that humans would let themselves become so unproductive simply because AI exists which can make so many of their decisions for them.
Even if the kind of super-intelligence he portrays could exist, I think it would be used to supplement human decision-making. It would do a lot of the work humans do not excel at or which is very dangerous or otherwise undesirable.
There would still be countless productive things humans could do. Including the creative arts, academic pursuits or anything else which the AI did was not quite as good as people. Which is actually quite a lot of things.
Which is why I do not greatly worry about things like robots taking jobs or the like. There will always be plenty of things for humans to do.
Then there is the fact that the human population is in many areas of the world, expected to seriously decline. Despite widespread paranoia about overpopulation. In this case, then there may very well not be nearly enough people to fill the jobs that need to be done to maintain an advanced civilization and we should welcome robots filling in for the shortage of people.
There is not really a great deal else to say about the prelude. It is largely intended as Tegmark presenting speculative fiction for the purpose of setting the scene for later chapters.
In this chapter, Tegmark defines life as a process which can retain its complexly and replicate. This is, of course, a drastically over-simplified definition of life that is not limited to biological life or any anything of the sort.
In fact, B to this definition, a purely mechanical system which manages to main at a constant (or improving) level of complexity and which is able to replicate itself is alive! So, that would mean that an assembly line able to create a copy of itself would be alive!
This is clearly a problematic and ridiculously oversimplified definition of life. But, Tegmark seems to have defined it so that it is not limited to biological life and can easily include anything that can replicate itself and maintain its complexity. This presumably makes it easier to include machines which meet little or none of the criteria for life used by biologists.
He then proceeds to divide life into three developmental stages. These three stages are:
Life 1.0: A biological stage where its hardware and software are evolved.
By hardware, he means the physical aspects of a life-form. In other words, its physical body. Its software is the information that life-form possesses.
This information would include memories, associations and the like. Lifeforms at this stage have no control over their software. Or in other words, they have no control over the contents of their brains and cannot “reprogram” it in any way.
Life 2.0: A cultural stage. At this stage, lifeforms have control over their software. That is, they can design their software by learning.
Life 3.0: A technological stage where a being can not only design its own software, but also its hardware.