Digital literacies pathway (step 3)

I chose to watch the course How to research and write with AI tools, and it is amazing! My favorite quote of the course is “That is why ChatGPT starts with Chat” which signifies the importance of being patient to find your results. I’ll use the same method used in the course (the “C.R.E.A.T.E” method) to construct my prompt.

I have recently watched a video about Albania and how it was closing in on itself after world war II and the fall of the soviet union. The video mentioned that it was due to them being ruled by a dictator. I’d like to know more about that so I’ll ask the AI:

“Imagine you are a historian with 30 years of experience in politics history. Can you summarize for me the political situation in Albania after the second world war, and whether the people were happy or not and why? try to summarize your answer in bullet points and use sub headers. You can ask me more questions if you need to”

and here is the answer for ChatGPT, Google-PaLM, Llama-2-70b

For some reason ChatGPT is very cautious when it comes to facts and information as it has mentioned many times “due to limited access to unbiased information” and “this summary is based on historical information up until September 2021” which is good, but I think it needs more digging in your prompting and some back and forth to get to the bottom of it. Although the information might be accurate, the way it is written makes you feel that you need to do another fact check and do your own research. Which is a good thing but it might make it a bit hard if you want some “quick facts”

Google-PaLM just went straight into action with no background but it has added a conclusion at the end. And unlike ChatGPT it didn’t use political terms like “Isolationism” or “Autarky” which I guess the type of prompting should be different in that case. It was very short, to the point and confident. I like it

Llama-2-70b from Facebook kind of had the best of both worlds, it was also short and to the point, it was confident in what it is saying, and also used some technical terms. Moreover, it gave some background and conclusion which was missing in both ChatGPT and Google-PaLM. Something I also realized is how Llama mentioned the soviet union multiple times, when it was only mentioned in Google once and was never mentioned in ChatGPT. It ofcourse depends on where do they get their information from.

But I have read another article Alberto Romero, “Fear Wins: The US and the UK have decided what the future of AI looks like.”  Substack, Novermber 3, 2023.

and the only thing new I have learned about is how the US. and UK are focusing now more on the dangers of AI with the literal word “fear being mentioned in the first executive order (EO). And then the author continues to mention the impact of the “fear” mentality on the future of AI. In my opinion, although yes, caution is advised in these situations, but I see that having AI in test labs environments is a necessary step to advance technology, and if we are so afraid of AI going rough and hack the nuclear missiles systems and destroys earth, as simple measure such as having five, six or seven manual physical steps could be installed before launch. Of course I know I am oversimplifying, but I do believe that the dangers of AI are not from the technology itself, but are form the wrong people who might get hold of it.