We are not prepared for GPT-4 and it’s time to talk about it.
The lightning-fast progress of OpenAI’s GPT models has undeniably revolutionized the world of artificial intelligence. However, this rapid evolution brings with it a set of concerns that we, as a society, must face head-on before fully embracing the powerful potential of GPT-4.
A first concern lies in our society’s unpreparedness to embrace such a potent tool without first educating people on how to handle this game-changing technology. For instance, many discussions revolve around the perils of providing GPT-4 access to high school students, who could potentially use it to complete all their homework and assignments.
This is being viewed as a danger rather than an opportunity to teach our kids the importance of studying essential topics and developing a critical sense instead of blindly copying answers from a magical tool to comply with the poor education systems.
Another concern is that the broad public does not understand how GPT transformer models work and can be easily fooled by the cleverness of this technology and believe everything that comes out of the model is true and correct.
A.I can make mistakes like all of us, perhaps that is the most human characteristic present in it, we all fail, but the rate at which A.I is advancing is such that its mistakes are getting far behind its assertions.
Theory of mind:
recent research shows that the current model of GPT-4 can keep track of the mental state in other people while evaluating the appropriate response, which means the A.I will take into consideration what it thinks you could be thinking while asking the question, this is incredible and must be acknowledged.
Final considerations
To finalize I would like to note that the most difficult problem with releasing GPT models to the public without training the broad public is that we ourselves have no way to define consciousness and therefore we have no way to test it.
Scientists don’t know for sure why transformer models work so well and thus we could be witnessing the rise of some kind of lower consciousness but there would be no way to tell, as we don’t understand it ourselves.
Disclaimer
I am not an A.I expert, I'm just a web developer who views the situation with eyes of worry and I believe it's important to talk about the subject as a group since these advancements are being released every few weeks we may not have that much time to react and adapt to the new systems in place.
Take what I wrote with a grain of salt and know that I could be wrong, nonetheless, I hope you found some value 👍🏽
There is another version of this article in Portuguese which can be found here:
