Unexplored Junior and LLM

@teramotodaiki: this year’sUnexploredJunior , I think the ChatGPT effect might raise the level of proposals! I have high expectations for this year’sUnexploredJunior. I think it is a very advanced skill to convey what you want to do to others in writing when you have not even fully organized what you want to do. Of course, what is important is the substance (content of the proposal), but there are many people who are losing out because of their writing skills. @teramotodaiki: everyone applying this year, please try ChatGPT as your wall-to-wall partner. There are many ways to use it. Hear how to write a proposal

Have them read the proposal and point out areas for improvement. I’ll just ask, “What do you think I should do to get in?” I’ll ask. …etc. I think it will shorten the time and improve the quality of the proposal. Maybe. ↓↓. @teramotodaiki: this is not my official position, but at least I am positive about using ChatGPT in proposals. I mean, if I were in your shoes, I would 100% use it… Rather, I’m a little concerned that some of you might be thinking, “Why not cheat like ChatGPT?” I am a little concerned that some of you may be thinking. At least the unexplored juniors are not prohibited.

@kaityo256: I like how you describe using ChatGPT for writing as “wallowing”.

• When I write a sentence in English, I plug the Japanese into DeepL, plug the resulting English sentence into Grammarly, and then check the corrected English sentence in Japanese using Google Translate.

@nishio: this is not the official view, but as to “whether it is ok to use”, I think “of course it is ok since it doesn’t say no”. I think the question that bothers me is “how is it beneficial to use”. My impression is that the output when used simply is about 75 points. In other words, if a score of 70 is the pass line, you’ll get it, but if it’s 80, you won’t.

• @nishio: If I say something like “submit 10 ideas that might be accepted for the Unexplored Junior,” I get a zillion ideas that look good at first glance, but I don’t think they will be accepted. 75 points won’t make you the best. Related story: How to come up with good ideas.

• @nishio: In school, it is OK to be “moderately good” because most people advance and only a few stay in school. But in a contest, only a few are selected and most are not, so the competitive environment is the opposite. In such an environment, it is not OK to produce a “moderately good” product. Average orientation is the worst environment.

• @nishio: If you ask an AI, “What should I do?” It may answer, “I should do this,” but there is no guarantee that the advice is correct. it’s not limited to AI. Often the advice of the adults around us is also wrong. ChatGPT is one of them.

• @nishio: it’s great to have an AI as a sounding board in that, unlike adults, they don’t bend over and get in a bad mood when you don’t follow their advice I think it’s a good idea to use AI as a sounding board. It never tells you the right way to do things, so you need to ask questions that may get the wrong answers back.

• @nishio: since anyone can use ChatGPT, that is, what you can do with ChatGPT, everyone else will do, so the competitive advantage It’s not a competitive advantage. I think the competition will be in what ChatGPT can’t do, such as your experience and enthusiasm for the project.

• @nishio: Let’s say, for example, you tried to make a prototype, but it was a huge failure and didn’t work out. This failure experience can only be detailed by those who actually acted on it. I think that episodes of actually acting and what happened as a result is an important weapon in a world where the starting line has been radically changed by ChatGPT.

mitoujr-mentoring-tips

relevance - ChatGPT writes book report

This page is auto-translated from /nishio/未踏ジュニアとLLM using DeepL. If you looks something interesting but the auto-translated English is not good enough to understand it, feel free to let me know at @nishio_en. I’m very happy to spread my thought to non-Japanese readers.