Yeah the example they gave does feel like pretty isolated unit test territory, or at least an integration test on a subset of the system that could be ran in isolation.

Go does have iterators: https://pkg.go.dev/iter

Thanks! Did not see this until your message, looking forward to make use of this

It would be a waste of time to develop right now, if it isn't on starship it would be a dead end in terms of progress. So they are better off just waiting until starship can be sent.

Was it actually a cut or was it not renewing something that was expiring? A bill to fund the government seems like the wrong place to be debating new spending.

I don’t know what the news rhetoric on all of this is, I haven’t seen it mentioned on here or on news articles, and I’m not on the socials/don’t watch tv. IIRC the initial ACA bill always had this cliff in order to make the numbers work for the bill to pass.

Like most long-term financial bills, everyone just assumed the cliff won’t hit and new legislation will pass.

This is the actual crux, no? “We expected the funny numbers to pass again” as opposed to “we should have addressed this before Biden left office”

AI PR reviews do end up providing useful comments. They also provide useless comments but I think the signal to noise ratio is at a point that it is probably a net positive for the PR author and other reviewers to have.

I liked the leaderboard prior to AI. Was fun to click through to various profiles and see who was in there and what their solutions looked like.

New Glen was meant to fly something around 6 times this year. At this point the best they will do is one additional launch to go with their first launch in January. Hard to see them doing any better timeline wise than SpaceX.

Training aside, an llm reading a pdf as part of a prompt feels similar to say Dropbox storing a pdf for you.

I bet you could get a court to say it was legally identical.

I think the Aereo case, and Scalia's dissent, are super relevant here. It's when the court decided to go with vibes, instead of facts. The inevitable result of that (which Scalia didn't predict) was selective enforcement.

edit: so what I really mean is that I bet you could get a court to say whatever you wanted about it if you were far wealthier and more influential than your opponents.

It's not similar at all because you can't get the book back out of the LLM like you can out of Dropbox. Copyright law is concerned with outputs, not with inputs. If you could make a machine that could create full exact copies of books without ever training on or copying those books, that would still be infringement.

> make a machine that could create full exact copies of books without ever training on or copying those books, that would still be infringement.

No it wouldn't. Making the machine is not making a copy of the book. Using the machine to make a copy of the book would be infringment because...you would be making a copy of the book.

Sure you could end up with occasional misinformation, but the speed at which you can get information more than makes up for it. Niche topics that would otherwise take hours or days to pull together and summar ise obscure sources takes minutes with LLMs.

I mean, yeah, if you don't particularly care whether the information you're getting has any grounding in reality.

Seems like a good benchmark for AGI. Start with things that are easy for humans but hard for LLMs currently.

But they have access to tools (though I'm not sure why they're not using them in this case).

Ask it to count using a coding tool, and it will always give you the right answer. Just as humans use tools to overcome their limits, LLMs should do the same.