Very good work for a high schooler, but explaining backpropagation and cross entropy loss is typical of ML paper-mill, though the feature extractor is interesting. These results really call for more domain-specific analysis to reach any significance deserving of headlines.

Because there is no admitting what was found were predictions. Millions of entities, that will take years to verify the data.

The interview is funny: when the winner was asked how he did it: I took that NASA database, and made the computer think...

No more concrete. Oh yes they said AI and infrared, he even used infrared.

All they know is his gut says maybe.

Several of the candidate variable objects are characterized in the results section of the paper. The model is also tested for effectiveness against synthetic data. It appears to be a useful method and the paper describes a plausible path for it to aid future discovery.

My assumption is always, a bright high school student has an impressive science fair project, but science reporting is terrible and misinterprets it as something more than it is.

(Also: "Kid outsmarts stuffy professionals" is an evergreen journalistic subject, and don't dismiss the political angle of sowing distrust in "establishment" scientists in favor of a younger person using AI)

Not that young people can't do big things but it's probably got less rigor than a graduate-level project.

Don't get me wrong, this is a really cool idea and it sounds like he did a great job. I don't want to be unjustly dismissive. These stories come up all the time and they usually don't amount to a whole lot- like most research.

Or, he goes to the polytechnic high school that’s right next to Caltech (half a block from the astronomy building no less) and getting research experience there is much easier than a regular high school.

Looks like he went to Pasadena High School though. When I did a bit of aerospace research at Caltech in high school all I did was cold email professors so any kid around here with some initiative and smarts can get connected.

I'm going through this now, a few years ago I heard that "research" with college professors was key - I ignored it because I thought it was performative and unlikely that any decent professor would want to waste their time. The kids that had their EDs accepted to Ivys at my daughter's school all did research with "professors" that were parents or friends of parents. It explains a lot of what I see of our graduates from "top" universities we hire from.

I live in a town with a few big pharmaceutical Co.s, and the local high school has several national science fair prizes (more than one would be very unlikely). All the kids that won prizes had parents working in R&D at those Co.s.

Sadly, I thought the same… Pasadena? Hmmm…

Regardless of whether there is something rotten here, I think they should in fact focus on the science and not the person behind the science. And that gives the young person some cover too.

The article says that The Astronomical Journal did just that: talked about the discovery without focusing on the age of the author. I think I prefer that.

Throw a $250,000 incentive into the mix and you're almost guaranteed to get less than honest work.

Whenever I see any front page OP I assume HN comments will dump on it, even one about a high schooler winning a science prize with a genuine scientific discovery. I hope you don't treat your own and your loved ones' news and accomplishments the same way.

Congratulations Matteo Paz. You not only won the science prize, you got the Hacker News front page treatment.

Or maybe it's the other way around?

I was an undergrad, not in high school, but I basically designed and ran a project that led to a major publication that a useless postdoc who was foisted onto the lab leader for political reasons took credit for an "kindly" gave me last author on.

Later, when I was in my own PhD, my adviser wouldn't let me continue MY OWN RESEARCH because it might "look like" I was stealing his work, so I had to invent a different thesis topic.

The above was in CS, this is more STEM BTW -- it's perfectly possible to develop a new computational technique to "discover" 1.5m new interesting things in an existing data set -- it's impressive, but it's not on par with say, a botanist cataloging 1.5 million new flowers.

My assumption is most Science Fairs are based on Dad and Mom’s contribution, and things like this are filler for elite university applications.

The one thing that suggests this might be legit is that Pasadena has an elite Math High School. https://www.mathacademy.us/

If a college senior could pull it off, perhaps a properly educated high school junior could too.

Mentorships are simply how most junior scientists get started. Even in grad school, most students initially take on projects their advisors have pre-qualified as interesting.

To make high school-level competitions more fair, we should likely prioritize access to researchers for all smart, hard-working high schoolers rather than only those who are nearby a university or have wealthy parents.

This is true of a lot of experiences in life though and isn't necessarily bad.

e.g. let's take a corporate example:

- New software is written to solve a problem

- It kind of works. At least, well enough that it's less of a problem

- An intern comes along and is told to make it better. They have nothing else to do so they give it their full attention for two months.

- Software runs 5x faster. Intern gets hired for doing such great work

Who should the credit for this? The person who originally solved the problem? or the intern who made it 5x faster?

At some point, does it matter? The original writer probably got credit for solving the problem and the intern got hired. Basically, everyone got some kind of benefit.

(This being HN, I am SURE there is going to be a debate about the above...)

Better than the postdoc I knew who was driving his PI's kids to football practice every day.

Do you have any evidence to back this up? I'm asking out of genuine curiosity

Yeah, it seems like so:

"I would like to acknowledge and thank deeply my mentor Davy (Dr. J. Davy Kirkpatrick) for introducing me to astronomy at IPAC and providing guidance throughout this project, aiding in data analysis and the collection of known objects for the test set."

https://iopscience.iop.org/article/10.3847/1538-3881/ad7fe6

this kind of cynicism hurts young people who actually do good work on their own though.

Totally agree. Most careers in Science are nepo since day zero.

This is spot on! Mostly, kids do less than nothing, their parents do the rest!

I would certainly believe this could be the case for this or the kind of science work that would be good for an application. Including this field.

There are of course probably fields where there is ~no grant money, thus barely any research. Einstein noted we only know .001% of what there is to note of the universe, and even then he was probably embellishing in the favor of knowledge.

I would also expect by the time you are a postdoc you are totally indoctrinated in your field in a way a high school student would not be. Standing on the shoulder of giants might not always be an advantage, if the giants have been whispering in your ear what to look at, whispering in your ear what they think is true, whispering in your ear what they think reality is, and all your fellows have been listening to whispers from similar giants.

I don’t know if the comment really adds to the conversation at all, but surely the comment should just be “kids are super smart”, if anything

From the paper [0]: "The computer used for this paper contains an NVIDIA Quadro RTX 6000 with 22 GB of VRAM, 200 GB of RAM, and a 32-core Xeon CPU, courtesy of Caltech."

That GPU was first released in 2018, and can be had for ~$1500 today. The computer as a whole sounds exactly in-line with what a lab would have as an old spare machine. The student is lucky for sure to have access to such an institution, but it's not like he had rich parents who casually handed him $10-$20k. Much more likely he got access to Caltech resources because his exceptional talent caused a professor to take interest in him:

"I would like to acknowledge and thank deeply my mentor Davy (Dr. J. Davy Kirkpatrick) for introducing me to astronomy at IPAC and providing guidance throughout this project, aiding in data analysis and the collection of known objects for the test set."

[0] https://iopscience.iop.org/article/10.3847/1538-3881/ad7fe6

You triggered an old memory of mine in high school of when I ran for class president in senior year and campaign spending was capped at $100 dollars and someone else flagrantly violated campaign finance rules and spent at least a thousand dollars primarily distributing pencils that would go on to litter the campus’ every corner.

Where is this number from?