Trying to Make Sense of Artificial Intelligence, ChatGPT and Dall-E
In this week's episode, Ricardo talks about ChatGPT again and comments on people's lack of knowledge and ability to assess how these artificial intelligence mechanisms work.
Many people believe that ChatGPT is a magic tool that can answer any question. However, behind the answers it provides, there is a complex network of algorithms designed to simulate human thought processes and assemble the answers from millions of pieces of information of all kinds, even including programming codes.
Neither ChatGPT nor Dall-E (which makes images from text) nor any other AI engine is perfect.
Ricardo has taken several tests over the last few days, getting correct answers and wrong answers.
Furthermore, the data that feeds these algorithms can easily contain cultural, social, or even, in extreme cases, racist bias.
The algorithm reflects real life. The answers are not necessarily free of bias.
Another alert is related to people's fear of losing their job or work because of artificial intelligence. In this case, Ricardo suggests a simple action: open a ChatGPT account and try it out. Meet and make your assessment.
Nobody knows for sure how artificial intelligence will be used in the future, but we can at least understand its current mechanism by doing our own tests.
By the way, the illustration of this episode was 100% created by artificial intelligence using OpenAI Dall-E (text: A Matrix Movie style photo of a person looking in a mirror and seeing a robot like the False Mirror by Rene Magritte)