Home Artificial Intelligence AutoGPT — The following big step in AI — Based on GPT-3.5 and GPT-4 — Almost AGI!

AutoGPT — The following big step in AI — Based on GPT-3.5 and GPT-4 — Almost AGI!

1
AutoGPT — The following big step in AI — Based on GPT-3.5 and GPT-4 — Almost AGI!

While browsing through the newest AI-news I stumbled upon a recent hot topic on github called “AutoGPT”. While I believed it will surely just be one other GPT-Plugin or Chrome-Extension, I soon realized that it was an entire lot more.

What’s AutoGPT and what can it do?

AutoGPT may thoroughly be the subsequent big step towards AGI that will revolutionize the way in which we work even greater than ChatGPT alone did.
In brief AutoGPT uses GPT-4 and GPT-3.5 via API to create full projects by iterating by itself prompts and by critically reviewing and constructing upon them in each iteration. It’s capable of read and write files, browse the net and review the outcomes of its own prompts, in addition to combining it with the said prompt history.

All you might have to do is give it an AI-Name — like “RecipeBuilder”, add an outline and 5 goals which the AI has to satisfy. After this is finished AutoGPT will start doing its own thing, until the project is complete.
As a bit disclaimer — I did realize that it was susceptible to fall into loops and make pointless requests if given too complicated tasks, but for easy jobs the outcomes are amazing.
Be warned that it is going to use credits out of your OpenAI-account, but there are 18$ included within the free version. Also AutoGPT will ask you for permission after every prompt, so you may test quite a bit before it is going to cost you a dollar.

For coding

I attempted doing a little easy coding with it, and if Python is used, the outcomes are pretty good (given the scripts aren’t too complicated). It struggled coding with other languages like Java in my case though. it’s capable of research data, put it into context after which make scripts with it, which is de facto astonishing in my standpoint. It might even do tests for its own scripts.

Non-coding

Much more interesting are the non-coding results, should you just let it research a subject and construct up knowledge to it and create the output for you. You may ask it to “make me wealthy” or to create full-fledged books with introduction and a fitting structure. For this text I’ll set AutoGPT up and let it create a recipe-book for the 20 best Japanese dishes, including explanations for exotic ingredients. You will discover the setup-part at the tip of the article called “set it up”.

Let’s create a Japanese recipe book

Here is my example of an app:
An AI that researches the perfect Japanese recipes online and creates a neighborhood textfile containing the 20 best Japanese recipes

Goal 1: Research the perfect 20 Japanese recipes
Goal 2: Create a recipe-book from it including an intro and outro for Japanese cuisine
Goal 3: give detailed information to exotic ingredients
Goal 4: give it a transparent structure
Goal 5: save the output to a neighborhood textfile in the present directory

It can start like this:

Note that it already critically reviews its own plans.

It then starts googling:

After getting loads of results, it starts to review the outcomes and comes up with a refined plan:

After browsing it is going to start to jot down into the file:

The file then looking something like this:

That is just crazy it created the recipe list, an inventory of exotic ingredients, an intro and commenced working. It added an increasing number of to the file in iterations and the list would have gotten larger and larger — full of information.

I did kill the applying (by entering “n” as a substitute of “y”) after a while, as I wanted to avoid wasting my credits (you might come back afterward though, because it has a inbuilt memory).

It’s really crazy what AutoGPT can do already, to not think what applications like it is going to find a way to do in some years and even months. It truly is value keeping track of the AI-news, I’m glad I’m here to witness AI making history.

As promised here’s the short guide to set it up — visit GitHub

Set it up

If you need to try the brand new hot stuff out for yourself, it isn’t very complicated, but you have to to get the needed API-Keys and in addition some very entry level “coding” should you can call it that basically. Be warned that it is going to use credits out of your OpenAI-account, but there are 18$ included within the free version.

  1. First, go to github.com and find AutoGPT.
  2. Install and register the necessities

3. Open your powershell or one other commandline-tool and enter:
git clone https://github.com/Torantulino/Auto-GPT.git

4. Then: cd 'Auto-GPT'

5. Finally: pip install -r requirements.txt

6. Navigate to your folder (On windows its current-user after which Auto-GPT) and rename: “.env.template” to “.env” and open it

7. Replace the Keys together with your API-Keys

8. In Powershell/CMD run: python scripts/predominant.py

9. Start constructing your project.

As mentioned it is going to ask you for permission before prompting, it’s also possible to give it permission to do multiple prompts in a row should you write “y -n” with “n” being the variety of permissions you’ll give.

Disclaimer: There may be also a dangerous “Chaos mode” (no asking for permission). Again — remember it uses credits out of your account. You’ll use this application at your personal risk.

Make sure — should you use it appropriately, you should have loads of fun with it.

More content at .

Enroll for our . Follow us on , , , and

? Try .

1 COMMENT

LEAVE A REPLY

Please enter your comment!
Please enter your name here