It’s becoming more and more obvious that we are at a very important time in human evolution. How will AI shape the future? We’ve had a number of ‘enhancement’ ages in our history when stuff got completely turned upside down and humans shot forward at new rates of productivity that completely changed how the world worked.Read more: How Will AI Shape the Future Of The Internet – Intelligence via LLM’s
At some point, we discovered how to create iron and steel. We discovered how to do agriculture on a massive scale. We had the industrial revolution. All these points in time slingshotted human evolution forward and really changed the way that the world worked. Now we are on the cusp of the AI revolution.
So everyone always thought that AI would go after the low-skilled jobs first. Why bother to do something hard like brain surgery when doing a low-skill job like making a hamburger could be automated in such an easy way? It made logical sense right? Only the thing we failed to see, was why to automate something that paid $10 an hour when you could automate something that pays thousands.
Big Brain Move
So it seems obvious now that a computer can completely destroy a human opponent in chess, but if we were to go back in history, this was deemed the pinnacle thing for AI, and something that everyone thought would be nearly impossible to achieve. Turns out a machine is actually incredibly good at looking at a problem from a lot of angles, thinking through all the possibilities, and ultimately seeing every possible outcome. Then it’s just a case of maths and looking through which is the most probable to make the win condition.
Actually, if we think about how good a machine is at looking at these many different possible outcomes and many different inputs, it suddenly becomes obvious why it is they would be able to handle these hugely complicated problems. A doctor is deemed to be an amazing doctor because he has seen and handled that same situation many times before. Whilst a good doctor might have performed that same operation 10 times every day for 10 years, a machine could see and learn from that many operations every minute. And that’s just the operations of that one doctor. The machine could see many examples from many doctors, reading in all the most minute details that a human might miss. This is when things get clever.
What version is your iPhone on?
So you probably have an iPhone or an Android phone, or perhaps use windows on your PC. If you think about version one of those tools, they are pretty laughable by today’s standards. But you don’t really rate your iPhone by version 1 right? You judge it by the version that’s in your pocket. Right now the big bit of tech that everyone is talking about is Chat GPT. It’s a cool bit of technology but it’s still only version 3. Your iPhone is on like version 14. Imagine what OpenAIs’ product might be like on version 14. We can’t even comprehend what that might be like.
If you think of image generators, they can already produce photorealistic images of almost any subject you can imagine. Want to see a picture of a cat riding a unicorn whilst sitting on top of the mars rover on Jupiter? No problem. And that’s only version 2! Pretty soon we will have videos being able to create like this too.
What could this mean?
So right now we haven’t seen any ‘obvious’ bad actors, but can you imagine what could come about? It’s so easy to produce AI images, it’s so easy to produce realistic-sounding content. It’s not a great stretch to imagine that countries might be motivated to control things like elections. Imagine how easy it would be with one billion dollars and the resources of a government to say, produce some images of a president leaving a whore house with a prostitute on each arm.
Now imagine how long it would take them to produce thousands upon thousands of news articles discussing the scandal and some different opinions on it, all from varied walks of life, languages, and backgrounds. It’s suddenly a much more scary thought and one that is very very powerful.