What did I learn from ChatGPT 3.5 on 1 week?
ChatGPT is a “Generative Pre-Trained Transformer.” “Generative” refers to an AI language model that can create new, original output, such as text, images, or other types of data that are similar in style and tone to the training data, but not identical. “Pre-Trained” means that the language model has been trained on large datasets, such as books and the internet. “Transformer” refers to the neural network architecture that is designed to process sequential data, such as text or speech.
It is REALLY good
Although there is a new ChatGPT 4.0, its predecessors work exceptionally well too. Thanks to their large datasets of knowledge, they can teach you numerous things. For example, I was trying to learn Python software from YouTube videos. However, ChatGPT is way more effective. It acts as a personal teacher that you can ask even the stupidest questions and get meaningful, reference-filled answers. If you did not understand a concept from earlier, you can ask, and it will give you a detailed answer. It’s like having a teacher who is full of knowledge and never gets bored.
I, as a scientist, always looking for references. I must search for the references it has. Where is this knowledge based on? ChatGPT is very good at that. It can give you the references for example wind energy usage percentage of USA in 2021.
It can do Grammar Check!
I have always relied on ad-filled grammar check sites, which require payment for extra grammar error corrections in the article. It must have been a shock for them to hear that ChatGPT provides free grammar checks. Actually, I don’t feel sorry for them. In Turkey, we are currently experiencing an inflation rate of 150%, yet these sites do not adjust their subscription fees according to nation by nation. This can make it difficult to write for people in high-inflation countries, and I expect something more from sites based in high-income countries.
I have always believed that technological research and improvements should be open access to everyone. I know that governments keep certain technologies secret from us, and that ChatGPT being open access does not change this fact. However, I cannot help but wonder about the technologies that the US government has access to, if ChatGPT is already this advanced. After using ChatGPT, I am more convinced than ever that all new technologies should be open access for everyone. Children in Hatay, for example, who do not currently have access to schools due to the earthquake, could benefit from ChatGPT as a personal teacher.
It is only one source.
This is what scares me about relying solely on ChatGPT for research. You are limited to just one source of information and have to accept what it provides. While you can always use Google to supplement your research, it has become more of an information dump than an information heaven. It’s important to search for peer-reviewed articles to ensure that you are getting the latest and most accurate information. To mitigate the risk of relying solely on ChatGPT 3.5, it is necessary to cross-check information with other sources and to be cautious about the information you obtain from it.
What the Future Holds?
There is no doubt that ChatGPT will have an important role in the future. In July 2019, Microsoft acquired OpenAI for 1 billion dollars. They are now introducing their “Copilot” features in Microsoft Office. With Copilot, you no longer have to take notes during meetings you cannot attend; it can handle that for you. It is not advisable to rely solely on C-level executives for insights on big datasets. “Copilot” will handle the insights from big datasets by inferring from patterns. We have seen large-scale layoffs at companies like Amazon, Google, Accenture and McKinsey. While the companies did not disclose their reasons for the layoffs, I have concerns that ChatGPT and Copilot may have played a role.
