Businesses Need Artificial “Specialist” Intelligence, Not the Other Kind

Here is the stark reality about businesses, we need specialists to do almost all our jobs. What this means in terms of the development of Artificial Intelligence (AI), or more precisely Artificial General Intelligence (AGI), is that the real world need of narrow intelligence outsizes the need for smarter self-aware intelligences. So the fear about the emergence of a sentient intelligence is completely misplaced. What we should really fear is the exponential pace of developing artificial specialist intelligence.

That is, the kind of intelligence that can do a job so damn well that we never need to pay a human anymore to perform the job. In the old days, the term “computer” was used for humans who would meticulously perform hand calculations. You would have rows of people that would do this kind of work:

http://www.computerhistory.org/revolution/calculators/1/65/272

In fact, you couldn’t hire just anybody off the street. These people had to be above average in their mathematical skills. Richard Feynman himself was known to be a very gifted “computer.” Here’s a story of Feynman besting the abacus where Feynman describes his approximation technique:

The number was 1729.03. I happened to know that a cubic foot contains 1728 cubic inches, so the answer is a tiny bit more than 12. The excess, 1.03 is only one part in nearly 2000, and I had learned in calculus that for small fractions, the cube root’s excess is one-third of the number’s excess. So all I had to do is find the fraction 1/1728, and multiply by 4 (divide by 3 and multiply by 12). So I was able to pull out a whole lot of digits that way.

So when Google has its self-driving cars drive two million miles, it has driven more miles than the average human. A human who drives on average 13,000 miles in a year will need to drive for 153 years to reach two million miles. After 1.7 miles, Google had reported that its cars had 11 minor accidents. I’ve driven my cars at one tenth the amount that Google has done and I’m certain to have at least 11 minor accidents (i.e scraped hubcaps, fender benders etc). The point though here is that automation can easily become the safest and most likely the best kind of driver we will ever have. There are 3.5 million truck drivers in the U.S., and pretty soon they will be 3.5 million truck drivers that drive less safely and less efficiently than specialized driving automation.

As far as the “street smarts” or “common sense” that we expect from humans to do their job correctly, we might as well throw that out the door also. Let’s look at the infamous case of United Airlines. A paying passenger was battered and dragged off a plane simply because common sense did not prevail. Surely all the flight attendants, ground crew and law enforcement were not complete idiots. Yet, a simple solution to the problem could not be found (i.e. raise the re-accommodation payment until someone accepts).

The reason we insert humans in our businesses is so that common sense will prevail. Unfortunately, the way we’ve mechanized most of our corporations, we know that we’ve thrown out all common sense. The popularity of the Dilbert cartoon strip is a testament of loss of common sense in our corporations. So let’s not pretend here, a majority of corporations run without much common-sense, furthermore they are run mostly by warm body humans. Clearly, the presence of a warm-body does not guarantee the existence of ‘common-sense’.

Most businesses don’t require their employees to have common-sense, what they want their employees to have is mechanized efficiency. It just turns out that this mechanized efficient kind of job that is the most easiest replaced with today’s technology (not some future AGI).

American workers are now forced into the “Gig Economy” whether they like it or not. Let me make it perfectly clear, the gig economy is a knee-jerk reaction to the destruction of jobs (or more precisely, that of careers). It is the survival mechanism for people to become more adaptable in what they can provide as services. What we are losing are single professions that are able to support ourselves with. It means that there are fewer and fewer jobs that we can support ourselves by being just specialists. So what happens is that people have to become specialists in many other kinds of jobs. That portfolio diversity allows us to survive as each job incrementally gets extinguished by automation.

So we are forced now to become generalists that become masters of many skills. How can we master many skills? It turns out that automation is what allows us to master skills with less experience. The table has been flipped in that folks with “common sense” and “street sense” become critical. Folks with advanced “business intuition” are the ones that will continue to thrive (before AGI). That’s because automation has given them superpowers (ask Andrew Ng).

So as we make rapid progress, expect to see more capable narrow intelligence applications created. The world will become even more competitive over time and anyone who doesn’t understand how to implement and deploy Artificial Intelligence is going to be at a complete disadvantage. Leaders of business who have an intuition of how their markets operate and have the savvy to build the automation that can exploit opportunities are going to take over the world.

The main point I want to make is that debates about the possibility of AGI (or AI if you still want to use that antiquated term) are pointless. The present reality (or is it clear and present danger?) is that artificial narrow intelligence is going to clean up big time. The main blind spot of many is that AI will only be useful if it has general intelligence. On the contrary, it is more useful if it has specialized intelligence.

This is not a new observation, Marc Andreesen has said many times that “software is eating the world.” If is not obvious to you, Artificial Intelligence is software.

Mark Cuban has an even more surprising quote:

Artificial Intelligence, deep learning, machine learning — whatever you’re doing if you don’t understand it — learn it. Because otherwise you’re going to be a dinosaur within 3 years.

♡ if you love brutal assessments.

Explore Deep Learning: Artificial Intuition: The Improbable Deep Learning Revolution
Exploit Deep Learning: The Deep Learning AI Playbook