What shifted my thinking over tim was the actual numbers. NASA's entire budget during the shuttle era was roughly 1% of federal spending [1]. We chose to de-institutionalize heathcare which really impacted homelessness. We didn't have to, but it was choice. And we could of done both. The failure was our leaders choosing not to, and that choice had nothing to do with NASA.

And the shuttle era, for all its problems, gave us Hubble. That single telescope showed us the universe is 13.8 billion years old, that expansion is accelerating, that nearly every galaxy has a supermassive black hole at its center. The shuttle crews serviced it five times to keep it running. I think it's hard to overstate what that one instrument did for our understanding of the universe.

I don't think the instinct you had as a kid was wrong at all! - And thanks for helping me re-activate some neurons- Whoopi made a real impession on me came from a real place. But I think we're lucky enough to live in a world where people fight to fix things on the ground and also point telescopes at the sky.

[1] https://en.wikipedia.org/wiki/Budget_of_NASA

After Year 1, WGA/SAG residual formulas decrease: Year 2: ~80% of Year 1 Year 3: ~55% Year 4+: sometimes stabilize at a “floor” rate

So what did they do? They ran it for a few years, ran the numbers, realized that Westworld was no longer profitable on the platform. (Profitable would have to mean draws enough new subscribers to the platform). AND THEN - Warner Bros. Discovery made new deals with other platforms with ads. I think you can still find Westworld on Tubi and other ad-supported platforms that actually pay Warner licensing fees.

It's pretty interesting how today's cars come with features like remote braking and monitoring cameras, all designed to make driving less demanding for us. So as these researchers work to make vehicles less distracting, these cool features somehow end up making us even more distracted. It's an ironic cycle that leaves you more distracted, and maybe more unsafe.

But now, like the OpenAI embedding you're talking about the embedding are constrained, trained for retrieval in mind. The pairs are ordered closer, easier to search.

I'm still fascinated at the prospect of these "star quakes" or magnetic flares that emerge from these stars. I guess these fields would weaken over time, but does it really maintain it's mass, just lose rotation speed or something?

  *He and his colleagues had also been engaged in a decades-long debate with a sprawling community of watchdogs — mostly doctors, lawyers and scientists from outside the agency — who were often broadly supportive of the agency’s mission but who fought with officials like Califf, sometimes bitterly, over the specifics: How should the F.D.A. be financed? What kind of evidence should new drugs and medical devices require? How should regulators weigh the concerns of industry against the needs of doctors, patients and consumers?*

Comments URL: https://news.ycombinator.com/item?id=43909057

Points: 3

# Comments: 0

[0] https://en.wikipedia.org/wiki/3Com_Audrey

Many doctorate holders contribute rich, well-researched content. However, if these editors predominantly come from similar demographic or cultural backgrounds, it may unintentionally lead to a less representative body of knowledge. I'm sure this kind of diversity is a common discussion and can be controversial to some folks.

[0] https://web.archive.org/web/20190724142049/https:/meta.wikim...

Comments URL: https://news.ycombinator.com/item?id=43101793

Points: 2

# Comments: 0