NASA-IBM Lunar Foundation open-Source Geospatial AI Model

newsroom.usra.edu

53 points·noobplus··6 comments

Good to see actual open-source models instead of these so-called "open weight" models which are still just inscrutable binary blobs that don't even provide a high-level catalog or really any description whatsoever of what went into the training data or the methods and tools used to train them.

It's always irked me. It's like saying a Photoshop binary is "open source" just because you can (theoretically) run it on your own computer instead of through a SaaS web app.

I mean if you push it, all software you can get your hands on is open source, it's just that binary code is a bit tougher to read than the form it was compiled from.

Also since you brought it up, I'd argue the difference between SaaS and local is way more important than "open source" vs "closed source". Closed source you can decompile. SaaS on someone else's cloud, you can do fuck all about.

That's why SaaS is so popular, too.

The word "source" explicitly means the pre-compiled code.

Dupe of https://news.ycombinator.com/item?id=49697549

“A rough guide for going back to the Moon” (ibm.com) - 162 comments

it is an open source from yesterday https://huggingface.co/nasa-ibm-ai4science/NASA-IBM-Lunar-Fo... these are collections https://huggingface.co/collections/nasa-ibm-ai4science/nasa-...

Reply on news.ycombinator.com