High school student uses AI to reveal 1.5 million previously unknown objects in space
https://phys.org/news/2025-04-high-school-student-ai-reveal.html
High school student uses AI to reveal 1.5 million previously unknown objects in space
https://phys.org/news/2025-04-high-school-student-ai-reveal.html
5 Comments
Impressive how an article about a remarkable use of “AI”, which was published in a peer-reviewed paper, gets downvoted simply because it uses the word “AI”. Based on this article, it seems like it’s a purpose-made machine learning program and not just using a LLM, so it’s not even the thing that people apparently hard-coded their brains to hate because people online said it’s bad.
Very commendable of him to do so and publish a paper on his results as well.
College students take notice as this may be one of the next check boxes to get into a high end college.
This is what AI is meant for, tedious tasks that would take far too much man power to be worth doing.
I read the comments here.
A few suggestions. Read the paper (not the article). Then comment and up or down vote.
Second – I think the great part of this is not the project or the paper, although I think it is a good start and it will improved upon. The great part of all of this is that should be emulated is getting more young people interested in science, earlier, and in a real scientific enviroment. This student is one to watch. Kudos to the people who helped and mentored this student.
Third. Some people are making a to-do about it being peer reviewed. The peer review process is very flawed. It needs to be fixed. The two issues that concern me the most is that many of the top scientists do not peer review. We need top scientists as part of the peer review process. Journals struggle to find peer reviewers. The quality of the peer viewers are not what they should/could be. Second – articles get both published that are not worthy and papers get denied that valid, cutting edge and in some cases repeatedly denied by reviewers. My favorite example is AI researcher Yann LeCunn’s repeated rejections of his AI research back in 2012 as being outdated, and “there is already is a better method for that” … HINT, spoiler alert – His rejected method is now the gold standard for image, pattern recognition that is used EVERYWHERE. TLDR; peer review is the best we have, but still throws away good science, and can push forward low quality science. You should read the science for yourself and decide. Do not say it is good science because it is peer reviewed. You be the judge.
Shitting on AI isn’t going to make it disappear, or somehow make your lack of a skillset more valuable. The salt on this is insane.