Some considerations after using ChatGPT for a few days.
1) It’s addictive: after a while you get used to asking him anything and even while you are working or studying you try to get help from him rather than wasting time doing tests, research and reasoning.
2) “Reasoning“: If you ask him complex questions, he can figure out what the crux of the matter is and come up with an articulate answer. He has synthesis skills, and if you ask him for clarification he elaborates on some aspects of the previous question and so on until he engages in a real discussion. He is consistent in his answers and corrects himself if he realizes he gave an inaccurate answer (almost always because you had asked the question ambiguously).
3) It will not be free forever: those who can afford access to these systems (professionals, companies or nations) will be at an incredible advantage over those who cannot.
4) It’s irritating: because it knows more than you do. It is useless, not even with ten degrees could you have the knowledge he has. And this is the beta version !!!
5) But how on earth is it possible that until recently neural networks did not distinguish dogs from cats and now they do these crazy things?
Sources:
- ChatGPT di OpenAI (AI for text production)
- DeepL (AI for text translation)
- Midjourney (AI for drawing)
- ChatGPT: crash course by Adrian Twarog.