The FAQ on the Jet announcement (https://typesafe.ai/blog/introducing-system-one-models-and-j...) claims it was trained with their own data:
replyWhere Does Our Training Data Come From?
TypeSafe is primarily a data research lab, which is how the biggest results in AI get made. We make all the data ourselves. We wouldn’t train on your data even if you asked us to (no offense). We do some pretty sophisticated stuff, but if you want to find out more, we’d have to hire you.
> We make all the data ourselves.
replyI mean, this claim is simply preposterous, and is discountable as ridiculous nonsense on its face.
Really? I read it as saying they use 100% synthetic data, which is made in house. Why is that preposterous?
replyYeah, and synthetic data is pretty darn good nowadays especially with access to frontier models.
reply> especially with access to frontier models
reply... so not made in house.
To create a cake from scratch, first you must create the universe.
replyHow do you create "100% synthetic data" that is filled with countless facts, coding patterns, medicine, law, philosophy, etc? The notion is farcical.
Sure, if your point is that nobody bakes cakes, you’re right.
replyWhile it is painfully, abundantly clear that such wasn't my point, if we need to stick to the cake analogy, it's like buying a cake from the grocery store, cutting it into six slices, and then reassembling the slices and now claiming that you made the cake yourself. It is, you see, "100% synthetic".
replyThis is a ridiculous conversation, but their claim is such laughable bullshit that it's amazing that anyone actually buys it.