What Are Autonomous Artificial Intelligence Agents? A Very Simple Explanation

To understand them, easily and very quickly, we have to go over how we got to this point.

So we started with this strange Auto-complete-like AI ability, where we could ask AI to say something of write a little program, in a chat.

Then really clever people curved this ability around it self, to make it reason, make the AI ask it self to make a little plan and check if the answer is correct.

People pampered these AI’s with extra instructions, and some created early agens, AI with a job, and maybe personality.

We ended up with smarter AI, still doing it things in a web page, and now brave people came around and brought these agents down to code.

Coding agents are equipped with a lot of if statements, and more text instructions, about code, patters, and programs they code.

Their if statements are in addition to their text instructions, the it-and-then’s structure their mind, remind the imperfect AI to check things.

When they reach the last if-then, they restart again, at the first if, re-ruining the cleverly made if-then procedure.

This is called a loop, but that is not a good name for it, they re-run the if statements, from the beginning, once they reach the end.

This worked really well for programmers, it is charming, impressive, the AI has a logical mind, and can pull up documents in the right if-then.

Almost just a couple of months ago, people went mad with this power, and they asked, well why does it need to be a coding agent.

And instead of AI being in a smart programming loop, they put it out in a normal operating system, and even scheduled tasks for it.

When you ask a coding agent to list files, it will list the files of the programming project.

When you ask an autonomous generic agent, it will list all the files on the computer.

This means, that it can run any command it is instructed to, if you make a command for posting products online, and give it a list.

It will automatically automate the entire process, schedule the creation of the list, giving it the steps you would take.

An now you have an autonomous product posting AI, that creates a lit of products and uploads list of products.

And then people went another step, this is very new, they asked the AI to improve the process.

To learn from what it goes through, and improve the procedures it follows.

They instructed the AI to generate memories that it can re-read, instincts it can create and track.

And to understand the power of agents, we have to recognize this as a kind of evolution of it’s text mind.

The autonomous agents are scheduled to check, to improve, to research, to test, and probably to think by subtle analogy to other things.

In a couple of months the world will take another step, and then another after that.

You have already fallen behind, needing to be told a friendly history of the thing to understand it.

Please, time has come for you to learn how to use Artificial Intelligence, to spend time thinking about Agents, and maybe testing some services.

If you can afford it, maybe run your own Autonomous one, the least expensive seems to run $600, you don’t have to connect it anywhere.

You have to learn, you have to stay with the times, you have to stay with it.

Post scriptum,

For those of you who are confused about the hardware, you need video ram 16 GB or more, I recommend but can’t afford 24GB.

I tried cleaning up an audio file with standard fast CPU, and I stopped it after 5 minutes.

With a graphics card, it only took 30 seconds, the difference between video card GPU and motherboard CPU is enormous.

Agents are standard computer programs wrapped around an AI, think of them as an AI harness, that strengthens the AI’s reasoning power.

An optimized AI is a 16GB or 24GB file that you load into your video memory, the files are called models, new one show up weekly, and are free to download.