In this episode, Jason Foster shares his thoughts on the new technology from OpenAI, ChatGPT, what it is, its origin, limitations and potential impact on humanity.
Listen to this episode on Spotify, iTunes, and Stitcher. You can also catch up on the previous episodes of the Hub & Spoken podcast when you subscribe.
For more on data, take a look at the webinars and events that we have lined up for you.
There has been a huge rise in the amount of AI available, but it is important to remember that it is still in its early stages. It can however still be of benefit if you want to experiment or gather new ideas.
00:10 The history of ChatGPT’s parent company OpenAI
00:53 What is ChatGPT and how it works
02:13 The limitations of ChatGPT
03:35 The ethical implications of AI
05:00 Comments on ChatGPT by its founder
At the moment, you may have seen a lot of discussion about different AI, from DALL-E creating artistic images through to trading bots in crypto and stock exchanges. However, none of these have been able to surpass the popularity of ChatGPT – a true AI that responds to native language. This means that instead of merely spitting out an input, it will sit and explain an outcome like a human would. People are even sitting down to have full conversations with ChatGPT. This shows that it is also capable of understanding complex conversational contexts so it can adjust its conversation style based on the user’s preferences.
ChatGPT is merely a chat AI so it lacks flow and finesse when it comes to certain tasks like creating a podcast which involves multiple steps such as script writing, audio mixing, editing, uploading, etc.
Because it uses neural networks to simulate human-like conversations, it can engage in conversation however it still has its limitations. The technology is based on natural language processing, so the responses generated tend to be grammatically correct but lack any kind of emotion or context. The system also struggles with complex topics and will often generate nonsensical or overly simplistic answers if asked something too vague. With Jason’s example, he demonstrated that the output quality is heavily reliant on what you instruct ChatGPT to do.
It is also unable to do physical tasks or comment on people, regularly citing difficulties when attempting to give subjective commentary on individuals or controversial events.
AI is exciting and it is progressing in leaps and bounds. It is not perfect yet, and there is still much work to be done to improve upon it. However, that is the beauty of AI – it has the ability to constantly improve itself with each new interaction users excitedly have with it.
And don’t rely too heavily on everything ChatGPT says. After all, it’s not perfect… yet.
P.S Did you see Jason’s plot twist coming? Comment and let us know on LinkedIn.