Google has announced a code red because they are a bit afraid of ChatGPT. Not that weird all things considered. ChatGTP, founded by Open AI, is the new kid on the block that is getting popular fast. Should we be happy with it or afraid of it? Is I, Robot a prediction or still fiction?
What is ChatGPT?
In short: An AI that can give you answers to all kinds of questions. You give it a question and it will reply in a way as if it looks a bit like someone is typing. Cool, huh? But not only questions work. You can also give it commands, as I will show you later in this article.
Who founded ChatGTP?
It is founded by Open AI, a startup company that already made some name. Maybe you have heard of GPT-3? (I didn’t, to be honest) GTP-3 is an AI that makes responses look like human-written responses. Another well-known product of Open AI is DALL-E, which is a bit more familiar… At least to me. This is an AI that can create images in the form of art or normal digital pictures.
And now Open AI has created ChatGPT. But who is Open AI? On their website:
“OpenAI is an AI research and deployment company. Our mission is to ensure that artificial general intelligence benefits all of humanity.“
Sounds a bit scary, if you ask me. If I read this I picture the SEO like this:

Still, I think this company, and especially ChatGPT, can change the world as we know it.
How can we use it?
Good question! I created an account at Open AI and tried ChatGPT myself. Let’s ask it some questions.
Let’s start simple and ask ChatGPT how I can check my memory with a Windows machine.

Okay, cool. A few suggestions that actually work. But I want to know how I can check my memory for errors. Let’s ask it in a different way:

That’s more like it. It gives a few options and the first one actually works because I tried it.
One of the things ChatGPT is really powerful for is developing software. You can ask ChatGPT to write a piece of code and it will spit it out. Let’s put it to the test:

And yes, this is the correct code. And as an extra, it also tells me how to use the AppendAllText method.
These are just a few simple examples. But ChatGPT can do way more. I was struggling with a plugin for WordPress. Or rather; how to make it. Google didn’t really give me the right examples and information. Mostly advertisements and a lot of results that didn’t come close to my search query. ChatGPT gave me the answer within 5 minutes.
Another example: I was trying to generate subtitles from a video I have recorded. Most software that could do that is expensive or online. Due to the fact we’re traveling, I need something offline. So I found Vosk. But the information about how to use it with C# is… Bad. I asked on StackOverflow and got a few standard replies:
- Show us the code!
Which code? That … that’s the whole point! - Have you used Google?
… Seriously? - Use software x, it’s online which is better and works very well with python
I asked for offline… and C#… - Show us the code!
Yes, twice…
And my favorite:
- This question is a duplicate of (link to a python code for a totally different software solution here). This thread is closed.
I went back to Google; but no luck. Then, after a week, I turned to ChatGPT. First I asked
How do I generate subtitles with Vosk using C#?
That gave me a lot of information about Vosk and even some hints why I kept getting errors. After reading it I could use the example code Vosk has been supplying.
Downsides
Yes, there are downsides to this great product. It can give wrong answers. So it’s really important to double-check the results you get. Don’t assume anything and use the information from ChatGTP to do some further research. ChatGPT is young and still learning and doesn’t have all the information yet. In some cases, ChatGPT is honest and says it doesn’t know.
As said before, ChatGPT uses information from different sources to learn. In some cases, it can learn the wrong thing. There have been reports about responses that were harmful to others.
More and more reports reach the surface about students using this kind of AI to write book reports, check the writing (grammar and more), and other things they should learn themselves. This could be a really big problem for schools.
These are the biggest problems for now. I am not talking about DALL-E, which has other problems on its own.
What do I believe about AI?
This a question I usually got from my students when we were talking about Azure Cloud AI. So this is my personal opinion:
I am a believer in AI, but in a way that AI does not really exist. I firmly believe AI can only exist as soon as it can have emotions. A part of an AI is that it should think for itself, and make decisions based on parameters and feelings.
Like with a child: If a child falls from a tree. That hurts (emotion). So next time the child will try not to fall or not climb the three, because it can hurt. Now, put the AI in a tree. An AI doesn’t feel pain, not have the capability to register the pain and learn from it. As long as no one – read: tech guy – tells the AI it hurts, he will keep falling, no matter what.
An AI is great with information. With certain parameters and algorithms, it can learn to recognize patterns and learn from them. If 8 out of 10 sources tell the Earth is round, and 1 says it’s flat, and 1 says it doesn’t know, the AI will assume the Earth is round,
Conclusion
Will I use it? For sure! Already doing that. If I hit a wall while developing a piece of code, I turn to ChatGPT and not Google. But I do double-check the results I get from ChatGPT.
ChatGTP is a next-level search engine. Not in a way we know from Google, Bing, or others. You ask and the AI delivers. Will it kick Google off its throne? I don’t know. At this point, we all need another search engine that shows results from other resources.
Yes, there are some problems with this system, but I think they can figure out a way that works for all of us.
What is your opinion? ChatGPT: Yay or nay?