AI Means "Planning" on Steroids

will it work?

One of the Things that all good people agreed on in the first half of the 20th century was “Planning.” You see it in John T. Flynn’s The Roosevelt Myth where Brain Trusters and others knew that the answer to the chaos of the Great Depression was Planning: “the Planned Economy and the theory of Spending and Debt”. And then, from Peter Hennessy’s Never Again, the British Labour government in 1945 would achieve “sustained growth… thanks to the efficiencies of a planned, mixed economy[.]”

I think we are seeing a revival of this notion in the current enthusiasm for Artificial Intelligence — which even President Trump is touting with his $500 billion Stargate program.

And of course don’t get me started on the NetZero plan to fight climate change, as though we have a clue what we are doing.

I believe The Science is quite clear on all this. I think it is best described by George Gilder in his book Knowledge and Power. You can read the short version from my blog here. Here is Gilder on page 215.

All information is surprise; only surprise qualifies as information. This is the fundamental axiom of information theory. Information is the change between what we knew before the transmission and what we know after it.

In other words, “planning” does not work because it is consciously trying to eliminate “surprise” and imagine that it can march to the future in predictable safety and without surprise.

Now, the election of President Trump — a “surprise!” — has brought to the fore a crew of “tech bros.” You saw them at the inauguration: Mark Zuckerberg, Sundar Pichai, Jeff Besos, Elon Musk & Co. And then there is the tech lord Marc Andreessen, the godfather of venture capitalism.

The whole point of venture capitalism is that you never know which tech startup is actually going to succeed. You sign up a bunch of startups, every one a real cool idea. My former neighbor across the street is CEO of a fusion power startup, name of Helion Energy. But will Helion Energy actually produce a fusion power station that works and power fleets of AI computers? We don’t know. I reduce it down to this meme:

There is a difference between a good idea and a good idea that works.

And the difference will surprise you.

Now, I am sure that AI is the best thing since sliced bread, but everything I have seen it do amounts to collecting information and organizing it. Can AI reproduce the startup culture that has been developed by tech lords like Marc Andreessen?

Can AI produce “surprise?” And if it could, would we want it to?

To me, AI looks suspiciously like “Planning,” the idea that we just have to sit a committee down in a room and “plan” the future. But whatabout “surprise?”

Here’s a surprise that just came in yesterday:

Scientists have confirmed the space around us appears to be growing faster than physics can explain, based off precise measurements of a galaxy cluster over 300 million light-years away.

So, do we just need to adjust a couple of coefficients, or do we need a major revision of our theory of the universe? See, my theory of knowledge is that it is all magic tricks and math that allow us to predict the movement of objects that we can observe. So, to me, it is not surprising that our present theories can’t explain the motion in “a galaxy cluster over 300 million light-years away.” I believe Kant’s dictum that we cannot know things-in-themselves, but only appearances, and all our knowledge is attempting to bridge the chasm between appearances and thing-in-themselves.

See, I believe that we humans completely underestimate how sophisticated we — and indeed all living things — already are, and that all our conscious knowledge is just icing on the cake. If we take Steve Jobs and the invention of the iPod, the forerunner of the iPhone, the question is: what exactly was involved in getting to the iPod? How much of it involved the curious combinations of traits that came together in Steve Jobs? How much of it was due to conscious intent by Jobs and his people, and how much due to unconscious instinct coded deep within our brains and DNA and accident?

On the other hand, maybe AI will do all the routine work of the world and leave us all free to come up with creative surprises to invent the future.

But then the question arises: would we want that, or would we just end up playing Minecraft with our friends and exchanging cat videos?