With AI, Astronomers Dig Up the Stars That Birthed the Milky Way

quantamagazine.org

14 pointspyman_in2 comments

The AI part: "They used an off-the-shelf algorithm called XGBoost, and trained it using high-quality spectral data from other surveys."

I imagine a funnier way, like an onion article where the astronomers just make pictures of galaxies with Midjourney V5 text-to-image, and their scientific discourse is arguing about prompts.

Well, if by off-the-shelf they mean kaggle winning open source ML library that's been around for like 10 years. This is the original goal of ML techniques applied to data though, not a meme prank. The magnification of higher quality output data from noisy input data in this example actually does make me wonder if we're not all that smart after all.

Reply on news.ycombinator.com