New Space Stations, Rocket Size Limit, Surviving Mass Extinction | Q&A Overtime 6

preview_player
Показать описание
Have I ever messed up a recording? Are there plans for new space stations? Is there a limit to the size of an orbital rocket? Will humanity survive the next mass extinction event? Answering all these questions and more in this week's Overtime Q&A.

🦄 Support us on Patreon:

📚 Suggest books in the book club:

00:00 Start
00:24 Can a software engineer help astronomers?
05:03 Do galaxies have ripples?
07:03 Can we build a telescope to study a specific planet?
09:14 Did I see the Apple ad?
15:15 Does AI hallucinate?
16:51 Is there a limit to the size of an orbital rocket?
17:57 Will humanity survive the next mass extinction event?
21:40 How would an electromagnetic cannon work?
24:56 Would Gaia work well at the orbit of Jupiter?
26:58 Will I get a new telescope?
28:10 Have I ever messed up a recording?
31:05 Are there plans for new space stations?

📰 EMAIL NEWSLETTER
Read by 70,000 people every Friday. Written by Fraser. No ads.

🎧 PODCASTS

🤳 OTHER SOCIAL MEDIA

📩 CONTACT FRASER

⚖️ LICENSE
Creative Commons Attribution 4.0 International (CC BY 4.0)

You are free to use my work for any purpose you like, just mention me as the source and link back to this video.
Рекомендации по теме
Комментарии
Автор

What they wanted to say: "We have compressed all those things into this one simple, wonderful device"
What we saw: "We have finally crushed the spirit of human creativity and now all you have is this bland, featureless device"

SomeoneExchangeable
Автор

I want to expand a little on Fraser's answer in 16:50, about the maximum size of a rocket:

A rocket's capability is defined by how much mass it can release and how fast it can release it, this is what defines the exhaust velocity or specific impulse (Isp), and is synonymous with how much thrust a rocket can generate with a given amount of propellant.

The problem is that there is a limit to how much Isp we can achieve with chemical propellants. Currently the record is 542 seconds, and that's using a propellant combination that is inherently impractical. For comparison, the RS-25 engines used on the Space Shuttle and SLS have a maximum Isp of 452 seconds. Theoretically we can go up to 1700 seconds with metallic hydrogen, but we don't even know how to mass produce metallic hydrogen, or even if it's metastable.

"But then just make it bigger." Okay, so let's consider a rocket that is 10m in diameter and 100m long. If we increase the length to 200m, the engines at the base of the first stage now have to lift 100m more of propellant (and dry mass), so they have to generate more thrust to lift the 100m of propellant we've added, and as we've seen, there's a limit to how much thrust we can get with a given amount of chemical propellant.

"So just add more engines and you're good to go." Well, if we increase the diameter of the rocket in order to add more engines at the base (and thus increase the thrust), we now have to fight not only the added mass, but also the Earth's atmosphere above the surface of the rocket, which will generate drag as we increase the speed, until we reach max-Q (the point of greatest aerodynamic pressure). You can increase the diameter from 10m to 20m, or even more, but at some point your rocket will face so much aerodynamic pressure that it will be unable to fight the atmosphere.

Add to this that the dry mass of the vehicle (propellant tanks, engines, pipes, etc.) are also facing Earth's gravity, so in order for them not to collapse you have to add more and more structural support, which will subtract from the payload your rocket can launch.

In theory if you were in a complete vacuum and weightlessness (basically in orbit) you could indefinitely increase the size of your rocket by making it wider and wider, since you would not face aerodynamic drag or structural limits.

Someone smarter than me could probably calculate an exact limit given some assumptions. My estimate is that building a rocket larger than 250 m tall and 50 m in diameter should be nearly impossible. Even if you used a multi-stage design, graphene for the main structure, and metallic hydrogen as propellant, I still don't think you could go much beyond those dimensions, unless you used antimatter.

Probably ~2000 t to orbit with chemical propellants is around the maximum payload limit you can get, and that's still 20 to 10 times higher than Starship.

nisenobody
Автор

Fraser Cain, week after week, month after month, you never cease to shine. Like the stars. Thank you.

ws
Автор

I am more annoyed that the AI and robot making companies seem to be focused on removing all of the fulfilling, fun, creative jobs humans can do first than I am with the tools themselves. I can have a computer make art of all kinds now, but it still doesn't make a burger, it still doesn't pick up the trash, it still doesn't grow the cocoa or the coffee or do any of the menial labor we were told it was supposed to do to make the lives of humans better.

peopleseethis
Автор

On "AI hallucinating" - the big thing to remember is that LLMs (Large Language Models, what most major "AI chat" things use) are *NOT* search engines. Some are supplemented with search engine like functionality to cut down on "hallucinating" when asked a simple question; but they do not have a set of "knowledge" that they try to compare your question to.

To use the websites example, they don't match "Nissan repair websites" in a database and give you the answer. What they're actually doing is they have a giant matrix of "segments of words" and what the "most likely next segment of a word" is.

So yes, in their giant list of data they have captured, the most likely next few bits of words to follow "Nissan repair website" may be things that do combine to be real websites, sometimes they won't. Sometimes the "most likely next bit" is nothing at all like a website name; or a wrong website name; or what looks like a valid website name but isn't.

"www." may be the most obvious next bit, so all 10 results will include that. Then "nis". But after item 7, what happens if in its data set it knows that Nissan used to be called Datsun? It may have enough to know that "nis" is next, but then have run out of "valid combinations" of "closest matches" that produce "san", and instead do "sun" because Datsun.

It's also why image generation AI (which isn't taught on words, but on images, ) often is *TERRIBLE* at adding text to images. Because it knows vaguely what letters look like in images, but it has no *language* context at all, so it just produces "things that look like letters". (Some newer ones do combine language and image, and can produce valid words in images.)

Machine Learning is an amazing tool with many great uses - but "generating factual blocks of text" is *NOT* one of its strong suits at the moment. (Finding planets among astronomical data sets _is_ something it should be good at, though.)

AnonymousFreakYT
Автор

But there is a bounce back problem. We’ve already gobbled up all the easy to reach minerals on the surface. Strip mined the next layer of easy-ish to reach as well. Sure it’ll all be in our landfills and ruins, ripe for the picking… except it’s all trapped inside plastic prisons. We all know how annoying excessive security packaging can be… Maybe humanity’s morlock descendants will finally learn the patience we never mastered.

olorin
Автор

As an artist apples add was a reminder i am both redundant and not necessary in the corporate world.

brianknow
Автор

Hey Fraser, I just want to say thank you for being an awesome person! You are intelligent, well-educated, enthusiastic about the things you find interesting, well-researched on the the things you talk about, level headed and logical on your opinions, and super kind, tolerant, and tactful when expressing your opinions out on the open wild internet. I often find myself agreeing with your opinions, and when I don't, I often find myself coming around to your side eventually because your arguments are logical and make sense. I really appreciate it. I want to thank your team too, because I'm sure it's a team effort and they are equally valuable. Keep up the great work, and thank you for making amazing content.

Madash
Автор

"Who needs payloads that big" sounds like Bill Gates when he said “640K ought to be enough for anybody.”

hawkdsl
Автор

I like how you put the Q&A and the ## in the bottle left of the thumbnail. Please keep doing that for all of the Q&A in the future. Thank you, either way, Fraser.

FBravo
Автор

I completely agree with you about iPads! We just cleaned out a sound booth at church and we found 4 useless iPads that had been left because they no longer worked. I started to look around for doors that might need to be propped open. Didn't find any so off to the trash they went.

BaryNusz
Автор

Fraser's recent interview about interstellar travel answered about max rocket size with on board propellant. I don't remember it exactly. Something like all the mass of the universe could only give you 1% of the speed of light? Because of the tyranny of the rocket equation, the issue is that it's so hard to get going when you have all that mass. That interview was about using energy provided externally of the spacecraft to reach speeds of 8% or even 24% the speed of light.

TheSkystrider
Автор

Fraser Cain's pontifications are my guilty pleasure lol

jbruso
Автор

Hallucination is a vocabulary word for AI. If you take any AI training for your job (if you haven't this is coming) then hallucination will be a word which means when the AI makes up things which aren't true.

JonnoPlays
Автор

The crumb in the mustache cracked me up 😂

alfriedrich
Автор

Never retire Fraser, we want to listen to you even when you go bald and grey (😮😅)
I’m watching this on an IPad (double strike 😂)
lol can’t wait for the livestreams to return 😁 enjoy the rest of your summer break 🎉

oldtimer
Автор

Mr. Cain, have you ever considered being a guest at DragonCon in Atlanta over Labor Day Weekend (Thursday-Monday)? There is a Space Track there that you'd be a huge it at, IMO.

MultiSteveB
Автор

@21:40 Lots of errors here: 1) The acceleration is a function of the length of the launcher, so you can launch humans too if you make the launcher long enough. 2) Rail guns can be fired multiple times, but again, you need to build a longer gentler accelerator than the navy guns, 3) You can launch into space without a circularization burn if you accelerate to escape velocity. That said, every launch technology needs a circularization burn to end up in a circular orbit - this is not unique to electromagnetic launchers.

pewterhacker
Автор

- Railguns can be used for orbital launches of extremely sturdy things like food, water, oxygen, and some spare parts. The electronics and thrusters within the capsule shouldn't be a problem as we currently have rocket-assisted artillery shells that endure much higher G-forces than a railgun would.
- Telescopes in the orbit of Jupiter would certainly allow us to refine the parallax measurements of nearby stars, and make estimates of much more distant stars.

pahtar
Автор

Fraser, Thank You four the awesome content and keep up!

TheAmental