Over the last few months, the internet has been buzzing about artificial intelligence, largely due to the release of ChatGPT. Lots of people have been having fun with it and lots of bad people have been making it do bad things. Even more people, those who fashion themselves as big-brained thought leaders, have been telling us what this means for the future of humanity. Of course, everyone selling a product is quickly adding the letters “AI” to their marketing kit.
The ChatGPT is an interesting project, but as the bad people referenced above have shown, it is still a work in progress. Even so, it does reveal two things that are important in the much larger discussion of artificial intelligence. One is that there has been some progress is creating what feels like intelligence. A bit of software that feels like it could be a human making small talk is a big step. That big step, however, is relative to where things were to this point.
This is a barely perceptible move forward, relative to where things need to be in order to start discussing anything close to artificial intelligence. In fact, you would need a much more powerful algorithm to detect the progress than what we currently possess, in order to quantify the progress. The main reason for that is we are not entirely sure if we can define human intelligence at the moment. In fact, we cannot even agree on where to start with such a project.
Now, HBD types and psychometrists might bristle at that, but crude methods to quantify aspects of intelligence are not the same as understanding the nature of it. We can sit two children down, give them a battery of intelligence tests and make some mostly accurate predictions about their life outcomes. We have enough data now to make statistically significant observations about relative intelligence. That is not the same as understanding the nature of human intelligence.
The reason for this is we do not understand human consciousness, of which intelligence is just one aspect. In fact, the human sciences are struggling to come up with a definition of consciousness. The physical sciences operate under the assumption that collections of non-conscious agents can come together somehow and create consciousness, but that is probably wrong. There is a good chance that it puts the cart before the horse.
That is, conscious agents are what cause non-conscious agents. What we think of as space-time and the objects within it are the result of human consciousness. This is the basis of Donald Hoffman’s book The Case Against Reality, in which he explains why much of what we assume to be true about the universe is probably wrong. For those who prefer an audio version of his work, here is a good interview of Hoffman by the ubiquitous YouTuber Lex Fridman.
Hoffman’s approach rests on two things that seem to be true. One is that nature does not reward accuracy in our perceptions, but rather the fitness of our perceptions, which Hoffman stated as Fitness Beats Truth. The short version is that evolution only cares about that which impacts fitness. As a result, a strategy that seeks only to improve fitness will, over time, dominate a strategy that seeks to create a more accurate perception of the world.
There is a lot of math involved in Hoffman’s presentation, but there is a much simpler way to think of it. We see in the marketplace for goods and services that the most popular product is never the ideal product. The Windows desktop won the computer wars not because it was the platonic ideal of the personal computer, but because it was good enough and cheap enough to win the fitness game relative to the other options, some of which were technically better.
The game of life works the same way. That which makes it easier for a living thing to replicate itself gets rewarded. That which inhibits replication gets punished. That is it and nothing more. If human perception evolved to improve our fitness, then it means it did not evolve to give us a more accurate perceptions of reality. That does not mean we live in a dream world. It just means that we cannot assume that our perception of non-contextual reality is accurate or even real.
The other thing Hoffman points out in his book is that theoretical physics has run into a very big problem. That problem is local realism. This is the principle of locality which states a thing is changed only if it is touched. The second half states that properties of objects are real and exist in our physical universe independent of our minds. This is the foundation of physics and it may not be true. Experiments have challenged locality and realism, suggesting one or both are false.
This brings us back to the artificial intelligence problem. The likelihood of us creating intelligence, much less conscious intelligence, even by accident is so low as to not be considered a serious topic of discussion. As Hoffman explains in his book, we are not close to understanding the nature of consciousness. We are not even sure the world we assume to be objective reality is more than an interface. It could literally be a figment of our collective imagination.
Like discussion of space aliens floating balloons over North America, talk of artificial intelligence is a nice distraction. We seem have a need to believe that such things are possible and perhaps discoverable in our time. In reality, they lie well outside our ability to grasp, other than the fictional. Intelligence is an aspect of consciousness and most likely that is not the result of non-conscious agents. That means intelligence is not simply clever math operating on non-conscious agents.
If a non-human intelligence were to visit our local concept of space-time from its place on the N-dimensional manifold that is reality, it would point out that our fancy new AI is not much more than a relatively faster calculator. From its perspective, it would be comically simplistic compared to our own consciousness. AI and the pursuit of it is a fun distraction, but it is not the dawn of super-intelligent robots enslaving humanity. We can do that without the aid of a clever chatbot.
If you like my work and wish to kick in a few bucks, you can buy me a beer. You can sign up for a SubscribeStar subscription and get some extra content. You can donate via PayPal. My crypto addresses are here for those who prefer that option. You can send gold bars to: Z Media LLC P.O. Box 432 Cockeysville, MD 21030-0432. Thank you for your support!
Promotions: We have a new addition to the list. Above Time Coffee Roasters are a small, dissident friendly company that makes coffee. They actually roast the beans themselves based on their own secret coffee magic. If you like coffee, buy it from these folks as they are great people who deserve your support.
Havamal Soap Works is the maker of natural, handmade soap and bath products. If you are looking to reduce the volume of man-made chemicals in your life, all-natural personal products are a good start. If you use this link you get 15% off of your purchase.
Minter & Richter Designs makes high-quality, hand-made by one guy in Boston, titanium wedding rings for men and women and they are now offering readers a fifteen percent discount on purchases if you use this link. If you are headed to Boston, they are also offering my readers 20% off their 5-star rated Airbnb. Just email them directly to book at sales@minterandrichterdesigns.com.