perhaps i must begin telling my ideas and stories in a video format, or at least in audio format. in the end of the article i will explain why exactly.
__
here i explain, still in writing, about my brief experience with using artificial intelligence tools. or rather testing them for a few days, in total of an hour or so. being late into the tests, regarding how long time the AI has been around, i did do a research into how AI works in principle, not to expect impossible and to know how to approach the technology.
please keep in mind that my analysis is based on very specific questions in science, in the fields that i have been interested in for a very long time, and know a thing or two on the topics which i asked opinions about from the AI, thus ..
what AI is good at:
pointing out errors and potential weaknesses in engineering approaches, based on well established science, reminding you about things you have forgotten to take into account while having an idea.
what AI is not good at:
understanding science which has advanced, discarding the previous understanding. the AI is referencing to the most frequently cited scientific data, taking it as a truth, while the research which has disproven the old data can be taken by AI as not a reliable information.
a conclusion:
the assistance by AI is valuable to quickly find errors in reasoning, but only applicable if you know the topic. you cannot rely on AI if you’re new to the field. at the present time, the AI is not capable of developing truly new understanding in science. humans can have an “aha” moment when evaluating the strength of factual data which may be contradicting, discarding the older data and deciding which data sets are not strong enough even if endlessly repeated. humans can decide what must be verified regardless of previous certainty in the absence of contradicting data, also taken into account the human nature of scientists, possibly producing fake data for personal benefits, and many more aspects of complex reality. the AI will rather calculate the truthfulness of data by those data points which are repeated the most, be it true or not. by the AI, those who shout the most and are the loudest, are right. quiet hard working scientists who point out discovered correct data and have no time to make a large promotion out of the discoveries, preferring to keep working, are ignored by the AI as irrelevant. in some cases AI is not even aware about important discoveries because a single data point, while being true, is not of a value against large set of the old but wrong knowledge found everywhere. the AI is no use for truly new science at this point. but it may become capable for it in the future, when the AI will be able to verify contradicting science not by the mass of the data points, but by the value of truthfulness in all available means of evaluation that humans have. currently AI is too specific in analysis, it needs to incorporate way larger set of understandings, including human nature and why humans create errors in data, be it by honest mistakes or on purpose, for personal pride, due to peer pressure, because of military secrets, to mislead business competition, and many other reasons. thus even analysis of large data sets, which AI can handle better than humans do, will not necessarily produce truthful information. a human must still verify the results in deciding the truthfulness and applicability in particular cases. wrong results may come not only from faulty data but also from errors in the way the data was taken for analysis, and the imperfect AI programs themselves. the AI remains as dumb as the programs created by humans not knowing deeply a particular field the AI is applied for. but as i said, AI is very useful in reminding you the things you may have overlooked, in case you know the subject well and understand that AI is correct in pointing out your mistakes. just believing the AI would lead to more trouble than having benefit from it. you still need to study a lot and not to rely on the AI in helping you out there where you are lacking in your knowledge.
the AI does will get better for sure, no question about it, but can you ever rely on it completely without human verification? if AI will arrive to a human level reasoning, becoming capable of correcting itself and finding flaws in its own reasoning, it may as well develop reasons to lie, the reasons only known to itself.
__
so, about writing. i’m not a native speaker of english. to express me correctly i must often verify the exact meaning of words, re-read the articles several times to make sure there’s a clear message in the sense i mean it, and so on. i’m not very articulate while talking though. i have spent years on learning, but very little time for expressing the ideas. you see, if someone reads my written articles with errors of expression in them, it may give an impression of not knowing the subject rather than being not a native speaker of the language. in contrast, when i talk, my accent and lack of vocabulary is obvious, so even when not being highly articulate it could make more sense to the listeners, who can in many cases understand from the context the word i was trying to tell, but had to replace it with another word i could find in my mind without seeking in a dictionary. it will save me time for studying new things, instead of spending the time on perfecting an article with several additions and corrections. currently i’m having a look into the latest advanced materials, discovered and created new properties, metamaterials and potential future construction materials, to see the viability of the project i mentioned in my latest articles, about the survival of civilisation.
telling my ideas and stories in audio/video is still just a project, but it may become a reality. regardless of living in a location with constant interruptions of silence, which is not ideal for audio recordings, the AI these days is already capable of cleaning up spoken audio from background noises. this is one application i could use the AI for, despite it being less capable in some other respects, especially in fundamental science.
__
to support the research and projects ..
https://old.rip/2022/12/26/about-contact-other-info/
__
new science 