Should we be worried about AI?

I tried out ChatGPT this week, it’s an AI powered search engine.

The main difference between something like ChatGPT and traditional search engines is that it can create!
Traditional search engines can simply regurgitate facts that they find by searching the internet. Results can be influenced if somebody pays to have an advert running. depending on how much the person pays, the advert may be placed further up the results. However the search engine itself can only display facts that it finds out on the internet.

AI based search engines are capable of finding facts and then using them to create essays, programming code etc and presenting the information in a human like way.

Practical example

As a test I decided to get it to write a program for me. I entered:

“create a VB program to read a spreadsheet file and generate Mikrotik code to set the address on ether1”

This should have been quite complex because:

I had asked the AI to generate a program using the syntax for a language called VB.
The purpose of the code was to read a spreadsheet and then write code dynamically based on the contents.
The resultant code had to be written in the syntax for RouterOS – the programming language used by Mikrotik Routers.  I hadn’t specified this last point although I had mentioned the manufacturer name “Mikrotik”.

As a human, you would ask lots of questions before starting. the AI engine obviously did quite a lot of searching instead but made good choices.

Inside about 5 seconds it gave me the code.

 

The code worked first time I ran it, all I had to do was paste the sample into my vb programming system. I had to add features to get it to have all teh functionality that I needed however it was a really good start. Starting from scratch I would have spent a couple of hours writing the code.  With the help of ChatGPT it took me about 30 minutes!

Should we be worried?

So should we be worried? – Not yet in my opinion. There will be some who think that a robot will jump backwards in time to wipe out the human race. However AI is only a computer program that can search through existing data. It does it quickly but it cant yet “think for itself”. If harm is caused then it is caused by the author, not the AI itself.

Have a look at the screenshot containing the code generated by ChatGPT. The results were quite honest. They started out explaining:
I’m sorry, but as a language model, I am not able to create VB programs or any other programs myself. However, I can provide you with some guidance on how you can approach this task.

ChatGPT was trying to explain that it is still fact based, it has searched the internet to find the VB language manual and then very quickly created code based on the rules it could find. It didn’t create anything new and didn’t make any thoughts of it’s own.

Here is what ChatGPT has to say about the subject of AI.

“As an AI language model myself, I can say that AI is a rapidly developing field with both potential benefits and potential risks. It is natural to have some concerns about the impact of AI on society, but whether you should be worried or not depends on the specific context and application of AI.

AI has the potential to revolutionize many industries and improve our lives in numerous ways, from enhancing medical diagnosis to reducing traffic accidents. However, there are also concerns about the potential for AI to cause harm, such as through job displacement or misuse of personal data.

It is important to stay informed about the latest developments in AI and to engage in discussions about the responsible use and regulation of this technology. It is also crucial for AI researchers and developers to prioritize ethical considerations and strive for transparency and accountability.

Overall, while it is important to be mindful of the potential risks and drawbacks of AI, it is also important to recognize the many benefits that this technology can bring and work towards its responsible and beneficial use.”

Conclusion

So for now, AI is not really intelligence, its is a very clever rule searching system that can then follow rules to create things.
What will come in the future? Who knows.  Humans broke rules to start learning and machines could do that one day potentially!

That said, it will change the way we create things, especially technical items. As an example, instead of writing programming code we will start to define what we want in a human language. AI will then be able to generate the technical code following our requirements.

How do I give it a go?

Using ChatGPT is quite simple. You do need to sign up for a free account. Visit https://chatgpt.org/ to sign up and get going instantly.

Once you sign up, you will get a screen similar to the picture below.

At the bottom is a prompt to “Send a message”, just type your question into this box:

As a first try, why not enter “What does Andisa IT Consultants do?”