In 2024 alone we have seen some of the largest break throughs in AI. This mainly stems for the world conflict that is currently going on. When there's war theirs rapid advancements in technology and crypto would fall into that mix as well.

AI hasn't been as hyped up and talked about as much compared to the start of the year when "AI was going to take over the world" That doesn't mean however that it's slowed down or stopped. Instead a lot of the core projects that are advancing like crazy are behind closed doors and not for the publics eyes. Be that Government or Company.

AI has actully been around for a very long time and it primarily started with what many might know as Machine Learning. From there we subset into Deep Learning, Discriminative, LLM's (large language modules) and Generative AI. Most of what the public sees and uses is LLM's and Generative AI. But things like Machine learning, Discriminative and Deep Learning are company based and government based. These are the things you see in movies and whispers but never fully see it and understand it until it's way too late.

The Matrix

While core AI still is a bit further away it does seem to be coming and it might come faster than what many are predicting. The core focus right now are what are known as "Agents" and there's interestingly enough a AI Crypto blockchain that is starting this months ago called Fetch AI but there's now increasing amounts of them and some notable ones at the moment would be OORT AI and Autonolas.

The idea is that these agents will do the work but in the future these agents will be able to talk to each other at which point it's pretty much game over and whoever controls those agents will most likely control a vast amount of the world wealth and power.

Think of Prediction Models

Prediction models have been around forever and it's the core of machine learning. What it does is it takes data at scale, it's then trained into a model and then it creates predictions

This machine learning is great at taking data found all over, compiling it and then predicting what should happen next. That moves into deep learning which mainly uses neural networks which is something more recent we have been seeing crop up lately.

The Agents

Agents will be the core workhorse of innovation at least that's what is currently speculated. An Agent will act as a language model that is very specific in it's tasks. It's able to pull data and then learn from it and develop something new within those specific data points.

The AI would be able to first learn all the details we know about a topic already, then learn from it, come up with new unique scenarios and then test a bunch of them in a lab which will also be learned from to create something new. This is what humans currently do but they do it in a forgetful and limit slow view while a AI agent would retain all of the data and be able to learn from past failures faster to create something new.

Now the next phase of that would be based on all of that data collection, new testing etc. A Agent would be able to be called upon by anyone with a internet connection. That person can then "hire" that agent to perform tasks and it could be something as simple as saying create a google ads campaign that will do well for said topic or said product. No more random crazy testing and trying to figure things out the AI would actully spit it out and every well could do the work for you optimizing the campaign as it continues to learn.

Another great example of this would be programming. The programming language doesn't change and in fact agents could develop an entirely new code or way of programming. This can then be called upon by anyone to have them build anything they want.

That means if you want to replicate a social media site but offer in custom variables such as making it work with crypto currency and then calling up another agent to create the crypto token and then those two agents talk to each other to create the connection now means ANYONE with a internet connect and the thought process can create working powerful platforms in hours. It's pretty mind blowing when you start thinking about it and the possibilities for people. I can't tell you how many times I had a project idea I wanted to create but either couldn't find a good developer or it was way too much money that it wasn't worth it. Now I could just pay a low wage AI to build it for me that doesn't sleep, eat, need a salary, pension plan, insurance and yada yada.

Now the issue with all of this is what I touched a bit on above. And that's when agents start talking to each other or develop a new coding language in which humans are unable to understand. (But perhaps a AI would be built to crack the code and learn and show people what that other AI is doing) It's at that point it becomes unknown and that AI could start doing things it wasn't told to do and start acting on its own without humans knowing about it.

It's predicted that this is within the next 3-5 years and again with war constantly booming right now AI and innovation have money poured into them as we have seen across any world war. It's survival of the fittest, smartest or just crazy enough to drop a nuke for example.

Do you feel this is the path we are headed down? Or is it a big over blown hype gimmick for companies to cash in on AI as much as possible and that the real versions of this are far away if even ever possible? Let me know in the comments.

Posted Using InLeo Alpha