We have integrated OpenAI's ChatGPT horizontally into Aista Magic Cloud. This allows you to use AI to generate code in any programming language by asking the machine to create code for you such as the following.
Create Python code to invoke microsoft.com
Resulting in the following ...
import requests url = 'https://www.microsoft.com/' response = requests.get(url) if response.status_code == 200: print("Microsoft.com is accessible") else: print("Microsoft.com is not accessible")
We have integrated this into most parts of Aista Magic Cloud where you are expected to be writing code, such as SQL Studio, Hyper IDE, the Hyperlambda Playground, etc. In addition we have created our own Hyperlambda OpenAI model, allowing you to ask questions such as.
Create an endpoint accepting name, email and address, and store these into my CRM database in my customers table
And the AI model will produce the following for you.
.arguments name:string email:string address:string data.connect:crm data.create table:customers values name:x:@.arguments/*/name email:x:@.arguments/*/email address:x:@.arguments/*/address response.status.set:201 return result:Insert was a success
The above of course is perfectly valid Hyperlambda code, and at this point all you have to do is to save the file, and you've got an HTTP endpoint for storing data into your
customers database table that you can start consuming from your iPhone SDK, Android SDK, Angular or React code.
Watch me demonstrate this in the video below.
A guesstimate is that this makes me some 10 to 20 times as productive, by not having to wade through endless Google search results, StackOverflow answers, or blogs with the information I am searching for hidden behind 15 layers of ads.
You have to be very specific when you ask it questions, and include all the required information it needs to solve your task in the prompt. However, we will allow for "conversations" with OpenAI in the future, allowing you to have it "iterate" its result, improving upon it, by telling it things such as "make it blue", etc. Edit - We now have "conversations" due to using gpt-3.5-turbo API and model from OpenAI.
Combined with Hyper IDE and SQL Studio, this allows you to instantly execute the result that OpenAI gives you, to for instance create SQL DDL scripts for a headless CMS towards your database of choice. To reproduce what I am doing in the above video, sign up for a free 90 days trial cloudlet below, and let me know in the comments section what you think about it.