Curious pug

use of AI chatbots for research

it appears like notebooklm, grok or gemini would be helpful in doing research especially parsing literature and identifying themes and trends. maybe even to make quick shortlists and screening of articles and abstracts.

the problem is that the AI chatbots are a bit of blackboxes at the user end and we never know how it shortlisted articles or screened them, despite providing comprehensive prompts. Further since these AI bots have context limitations, we dont know how much of the input information they are truncating.

Further the articles are written by humans for humans, such that a quick scan of a paper provides much more intuition of its contents than asking an AI chat bot to summarize it for you. the summary that is generated is based on a generic prompt pre-programmed by the makers of the AI chat bots. therefore the summaries dont have the depth that a good author puts into the article writing.

i have just begun scanning articles after 9 months of using AI to summarize things for me, and the clarity im receiving in just 1 hr of scanning articles is much higher than any of the garble that has come out of the AI bots. Out mind just doesnt internalize what the bots say. Your memory is affected and your critical thinking is affected.

Hence when conducting research, AI chat bots impede your learning - which is the development of your memory and critical thinking ability.