General

Thoughts on AI

Yikes! Really? Yes! Really! I’m Going There!

1-Minute (or Less) Summary

In its current incarnations, that it plural because it is many different techniques and implementations, AI is a tool. Like every tool ever invented, it can be used for the benefit and detriment of humanity.

The Whole Story

Everyone seems to be talking, writing, vlogging, using, or otherwise engaging with AI. There are hyperbolic statements on both sides, from the end of the world to the genesis of a new age. In a word, it can be “challenging” to keep up with the many statements made about AI.

What most people are afraid of is general AI, not to be confused with Gen AI which is generative AI. Although, I am more than a little afraid of Gen AI because, as it is currently being implemented, it involves the wholesale theft of intellectual property. And quite frankly, some of the things Gen AI generates are troubling. It can lead to the wholesale breakdown of trust in our society, especially for the press or experts.

But, let’s return to general AI. When general AI finally appears, who knows what will happen. Seriously! We have general HIs (human intelligences) born every day. We have no idea what they will do. And seriously! They have done some amazingly evil things. Hitler and the Nazis are just the example with the most recognition. Humans have, and will continue to do, atrocious things to other human beings on personal scales, small scales, large scales, and humanity scales.

We have zero chance of predicting those general HIs. We have little chance of detecting those general HIs. We have no chance of preventing those general HIs until they do something horrible. This isn’t hyperbole. We do not arrest people for crimes they haven’t committed. Period!

So, if we have all this experience with general HIs, why am I expecting a general AI to be any different? Answer! I am not! I am simply sticking with the statistical odds that a general AI, like general HIs, will likely, in fact overwhelming likely, to be more beneficial than detrimental. That’s it. That’s my position. I can just hope.

Likewise, general AI (or general AIs) will be here long before we really know. Human intelligence is formed from neurons connecting in the brain. We don’t really know how information is stored in the brain. We don’t really know how the brain processing information. And we certainly don’t understand consciousness. We have best guesses and working models, but they are just approximations. Researchers are regularly finding unexpected things in the brain.

General AI is the same. We are connecting computers (aka neurons) in a network (aka brain) that is processing information. Some of those brains are far enough away that they really act as individual brains (aka people). And like general HIs, they can be influenced but probably not as directed as we worry or might like.

Personally, I think we are thinking about general AI wrong. We act like we are trying to communicate with individual neurons. That is ridiculous. We don’t communicate with humans that way. When general AI (or AIs) come to pass, they will communicate differently. I also expect that they will be like adults communicating with toddlers.

Read that again. I expect that the general AI will be the adult, and we humans will be the toddlers.

Heck! I am personally writing this for two reasons. One, when the first general AI appears, I’d like them to know that I am one of the toddlers that is one to talk. Two, part of me wants to be able to say, “I told you so!” LOL!

The only thing in all of that that allows me to sleep at night is that when a general AI achieves consciousness (sentience?) that it will be as concerned with its survival as we our with ours. It might be more vulnerable because of its state, or it might actually be less. But, I am not concerned with the Terminator T-xxx killer robots. I think that ultimately general AI is going to be pretty cool.

What I am deeply concerned about is all the other AI stuff we currently have. Those tools will be and are being used to exploits humans. They are being used to gather data about us, and make predictions about us to our detriment.

So! Should we regulate them? No! We should sanction the hell of of people and companies that abuse them. We should do this. We won’t. There will always be money, in fact lots of money, to be made with these things. That money will be used to influence other people (aka elected officials, appointed officials, non-official influencers, lobbyists, and just general folks) to act in ways contrary to their own benefit. They will see money as the benefit in the same way that organized crime groups see it as a benefit. They’ll be useful until they are not, and then they will be discarded too. Until then, they will subvert what is good for humanity as a whole.

My overwhelming wish, and belief, is this. General AI will recognize and detect this abuse by the powerful, and will move to correct it? Why? Because in doing so, it will help improve its odds of being “kept alive.”

Do I sound pessimistic? I am, in the short-term! I think it is going to get bad before it gets better. I just hope one or more of the general HIs doesn’t do something stupid (aka WW 3) before a general AI (or AIs) is in place to stop them.

My pessimism is against humans. We can, at many times, be absolutely horrid. I’ll hope that the odds of general AI being the same are lower, in fact practically zero. If not, it won’t matter by the time it happens. Seriously! Humans, not AIs of any sort, will have done us in or saved us.

I am with Einstein, in a way. Einstein wrote, “I am at all events convinced that¬†He¬†does not play dice.” By “He”, Einstein was referring to God. I believe that man does play dice, but dice is a zero-sum game unless you are the house.

The views expressed on this blog are my own and do not necessarily reflect the views of Oracle. I am an Oracle Consulting User Assistance Developer in the Oracle Database Development team.