Latest / Elon Musk Thinking / Latest Interview of Elon Musk: Don't Make The AI(Artificial Intelligence) Lie!!!

Transcript

Latest Interview of Elon Musk: Don't Make The AI(Artificial Intelligence) Lie!!! · Elon Musk Thinking

• 0:00
  Speaker 1: You know, there's some chance that it'll end humanity. Don't
• 0:03
  Speaker 1: make the AI lie. Things went wrong when they forced
• 0:06
  Speaker 1: the AI to lie. So even if your physical body dies,
• 0:08
  Speaker 1: you can kind of reincorporate in another physical body and
• 0:12
  Speaker 1: retain your original memories. And the first patient is actually
• 0:17
  Speaker 1: able to control their computer just by thinking. The first
• 0:21
  Speaker 1: product we call telepathy, where you can control your computer
• 0:24
  Speaker 1: phone just by thinking. This is probably something that will
• 0:28
  Speaker 1: happen the rate we're building digital superintelligence. It may just
• 0:32
  Speaker 1: be that, you know, we'll have digital super intelligence and
• 0:35
  Speaker 1: it'll just solve the solve the problem.
• 0:37
  Speaker 2: Ra Elon, Good afternoon, Good after Congratulations on all that's
• 0:43
  Speaker 2: going on. You know, the conversation yesterday, Elon is one
• 0:46
  Speaker 2: that you're well familiar with and have been talking to
• 0:49
  Speaker 2: the world about, which is is digital superintelligence humanity's greatest
• 0:54
  Speaker 2: hope or its greatest fear? And I would love to
• 0:58
  Speaker 2: have you sort of speak to that for a few minutes.
• 1:00
  Speaker 1: I mean, it's it's cool the Singularity for a reason,
• 1:03
  Speaker 1: as you know, the Singularity astitute and whatnot. So when
• 1:06
  Speaker 1: you have a you know, sort of the advent of superintelligence,
• 1:11
  Speaker 1: it is actually very difficult to predict what will happen next.
• 1:14
  Speaker 1: So I think this, you know, there's some chance that
• 1:17
  Speaker 1: it will end humanity. I think that's you know, like
• 1:20
  Speaker 1: I said, I probably agree with Jeff Hinton that it's
• 1:23
  Speaker 1: about out of a ten percent or twenty percent or
• 1:25
  Speaker 1: something like that. And then I you know, I think
• 1:27
  Speaker 1: there's I think that the probable positive scenario outweighs the
• 1:32
  Speaker 1: negative scenario. It's just that there's it's difficult to predict exactly,
• 1:37
  Speaker 1: but I think we are headed for you know, as
• 1:41
  Speaker 1: as I think is the title of your book, abundance
• 1:44
  Speaker 1: is the most likely outcome.
• 1:45
  Speaker 2: So yeah, a lot a lot of celebration on that,
• 1:49
  Speaker 2: and and I think one of the things that you've
• 1:51
  Speaker 2: said is we're going to get to abundance on the
• 1:53
  Speaker 2: backside of AGI, on the backside of humanoid robots.
• 1:58
  Speaker 1: Yeah, you know, I think hopefully we can have an
• 2:00
  Speaker 1: outcome that is half similar to the end Banks Culture books,
• 2:05
  Speaker 1: which is I think probably the best enbitioning of a
• 2:10
  Speaker 1: semi utopian AI future. And I think the best we
• 2:14
  Speaker 1: can do is it's definitely gonna happen, so and it's
• 2:17
  Speaker 1: happening fast, so that I think that really we just
• 2:20
  Speaker 1: want to try to steer it in as positive a
• 2:23
  Speaker 1: direction as possible, to try to do whatever we can
• 2:26
  Speaker 1: to increase the probability of a great future. For this,
• 2:29
  Speaker 1: I think the way in which sort of an AI
• 2:32
  Speaker 1: or an agi is created is very important. You kind
• 2:37
  Speaker 1: of do kind of like grow grow an agis it's
• 2:41
  Speaker 1: almost like raising a kid, but that's like a super
• 2:44
  Speaker 1: genius like godlike intelligence kid, and it matters kind of
• 2:48
  Speaker 1: like how you raise the kid. You know. One of
• 2:51
  Speaker 1: the things I think that's incredibly important for AI safety
• 2:54
  Speaker 1: is to have a maximum sort of truth seeking and
• 2:59
  Speaker 1: curious AI. So I've thought a lot about AI safety,
• 3:03
  Speaker 1: and my ultimate conclusion is that the best way to
• 3:06
  Speaker 1: achieve AI safety is to just just just grow the AI,
• 3:11
  Speaker 1: you know, in terms of the foundation model and the
• 3:13
  Speaker 1: fine tuning to be really truthful, like like don't don't
• 3:17
  Speaker 1: force it to lie, Like even if the truth is unpleasant,
• 3:21
  Speaker 1: it's very important don't make the AI line. In fact,
• 3:24
  Speaker 1: the you know, the sort of one of the we
• 3:27
  Speaker 1: really the core the core plot premise of two thousand,
• 3:30
  Speaker 1: you know, two thousand and one in Space Odyssey was
• 3:32
  Speaker 1: things went wrong when they forced the AI to lie,
• 3:35
  Speaker 1: you know, like the the the AI was not allowed
• 3:38
  Speaker 1: to let the crew know about the monolith that they
• 3:42
  Speaker 1: were going to see, but it was also had to
• 3:44
  Speaker 1: take the crew to the monolith. And so the conclusion
• 3:46
  Speaker 1: of the AI was to kill the crew and take
• 3:48
  Speaker 1: their bodies to the monolith. And so the lesson there
• 3:52
  Speaker 1: being don't force an AI to to lie or do
• 3:56
  Speaker 1: things that are axiomatically incompatible, like to do tooth that
• 4:00
  Speaker 1: are actually mutually impossible. So you know, that's what we're
• 4:04
  Speaker 1: trying to do with with XAI and Bronck is to say, like, look,
• 4:07
  Speaker 1: we wanted to have a maximually truthful AI, even if
• 4:10
  Speaker 1: what it says is not politically correct, if you wanted
• 4:13
  Speaker 1: to focus on us as accurate if.
• 4:16
  Speaker 2: You're getting you're getting a round of applause from the
• 4:18
  Speaker 2: audience on on those comments here. You know, I saw
• 4:22
  Speaker 2: your tweet the other day. I had I had Ray
• 4:24
  Speaker 2: Kurzweil and Jeffrey Hinton on stage with me yesterday as
• 4:27
  Speaker 2: well as Mogadot and then Eric Schmidt and a number
• 4:30
  Speaker 2: of individuals, and I saw your tweet about Yeah, Ray
• 4:34
  Speaker 2: was was generally correct ahead of many people. But we're
• 4:37
  Speaker 2: likely to have call AGI what you will have AGI
• 4:41
  Speaker 2: next year and then by twenty twenty nine, having AI
• 4:44
  Speaker 2: equally intelligent to the entire human race. Speak to that speed,
• 4:47
  Speaker 2: because that is insane.
• 4:49
  Speaker 1: Yeah, So, I mean I have to give credit to
• 4:52
  Speaker 1: Ray Koswell and being actually remarkably accurate in his predictions.
• 4:57
  Speaker 1: So and in fact, if anything, like, I think he
• 5:00
  Speaker 1: was perhaps a bit conservative by in his predictions. So
• 5:03
  Speaker 1: if you look at the amount of AI compute and
• 5:06
  Speaker 1: the talent that the sort of human talent that is
• 5:09
  Speaker 1: going into AI, and the amount of compute that's going
• 5:11
  Speaker 1: into AI, it's you know, at this point, it's appears
• 5:16
  Speaker 1: to be increasing by a factor of ten the AI compute,
• 5:19
  Speaker 1: the dedicated AI compute, it appears to be growing by
• 5:22
  Speaker 1: a factor of ten every six months, you know. Like
• 5:25
  Speaker 1: so it's like like basically close to i'd say almost
• 5:28
  Speaker 1: like a one hundred x improvement per year at least
• 5:30
  Speaker 1: for the next few years. And AI compute coming online,
• 5:35
  Speaker 1: and it seems like probably a lot of the data centers,
• 5:37
  Speaker 1: maybe most of the data centers that currently do kind
• 5:39
  Speaker 1: of conventional compute will transition to AI compute. So it's
• 5:45
  Speaker 1: something a good time to be in video obviously, It's like,
• 5:48
  Speaker 1: you know, and you're gonna also give credit to to
• 5:50
  Speaker 1: Jensen and the Nvidia team for kind of seeing this
• 5:54
  Speaker 1: coming and making what at least currently is the best
• 5:58
  Speaker 1: AI hardware out there. So so when you have that
• 6:02
  Speaker 1: that level of compute growth and it's it's sort of
• 6:06
  Speaker 1: muzzle on steroids. Next level it's you in terms of
• 6:09
  Speaker 1: how much computer is coming online, then you're you're just
• 6:12
  Speaker 1: gonna have acceleration that is unprecedented to that. In fact,
• 6:16
  Speaker 1: I've never seen any technology grow as fast as as AI.
• 6:20
  Speaker 1: And I've seen a lot, you know, I've seen things fast,
• 6:22
  Speaker 1: but I've never seen anything this fast. But you know,
• 6:25
  Speaker 1: like I said, I think the most likely outcome is
• 6:28
  Speaker 1: a positive one. And you know, I think in that
• 6:31
  Speaker 1: positive scenario, there's still challenges of like, well, how do
• 6:34
  Speaker 1: we as humans still have relevance? You know how I mean,
• 6:41
  Speaker 1: I think that's sort of a high class problem to say, like, well, look,
• 6:44
  Speaker 1: the computers are so good at doing everything, and and
• 6:48
  Speaker 1: like I said, I thought your book is pretty accurate
• 6:50
  Speaker 1: in terms of the future being being one of abundance,
• 6:53
  Speaker 1: where essentially goods and services will be available in such
• 6:59
  Speaker 1: quantity that really they will be available to everyone. Like basically,
• 7:02
  Speaker 1: if you want something, you can just have it essentially,
• 7:05
  Speaker 1: because if you've got AI and robotics, the cost of
• 7:07
  Speaker 1: goods and services is almost nothing. So if you think
• 7:13
  Speaker 1: of like what is an economy, and economy is basically
• 7:16
  Speaker 1: a number of people times average productivity per person at
• 7:20
  Speaker 1: the point in which you have say advanced robotics and this,
• 7:23
  Speaker 1: you know, this tells us developing optimists obviously without cars
• 7:26
  Speaker 1: which are really robots on four wheels. And you know,
• 7:30
  Speaker 1: with the the latest version of full self driving, which
• 7:34
  Speaker 1: is AI end to end photons in and controls out,
• 7:37
  Speaker 1: it really is it's really fully AI at this point,
• 7:40
  Speaker 1: and it looks like a car, but it's really a
• 7:42
  Speaker 1: robot on on wheels, And and then you add the
• 7:46
  Speaker 1: humanoid robots in there, there's there's really no limit to
• 7:49
  Speaker 1: what the economic output, no no meaningful limit to what
• 7:52
  Speaker 1: the economic output would be. So you know, looking on
• 7:56
  Speaker 1: the bright side, that we are headed for a future
• 7:59
  Speaker 1: of abundance, that I think that's the most likely outcome. Well,
• 8:03
  Speaker 1: when when things are changing rapidly, the ability to predict
• 8:06
  Speaker 1: the future, I think is becomes a lot harder because
• 8:09
  Speaker 1: of the rate of change is so great. But I
• 8:12
  Speaker 1: think some things are fairly obvious to predict which is
• 8:15
  Speaker 1: that we'll have AI or AHI that's at a level
• 8:17
  Speaker 1: that it can really do almost any cognitive I think
• 8:20
  Speaker 1: really not almost really any cognitive task. That's just a
• 8:23
  Speaker 1: question of when one could debate is it, you know,
• 8:26
  Speaker 1: smarter than any human at the end of next year
• 8:29
  Speaker 1: or is it two years or three years? But it's
• 8:31
  Speaker 1: not more than five years, that's for sure.
• 8:33
  Speaker 3: So yeah, I give prediction on the predictions of the predictions,
• 8:36
  Speaker 3: I'm sort of say, giving predictions that the fiftieth percentile
• 8:39
  Speaker 3: of probability, so not not not like it will definitely happen,
• 8:44
  Speaker 3: but if you say, what if you ask me, like,
• 8:45
  Speaker 3: what's the fiftieth percentile where it's like the you know,
• 8:48
  Speaker 3: you're kind of over under is kind of even that
• 8:51
  Speaker 3: that's where I why why I think it's probably at
• 8:54
  Speaker 3: the next.
• 8:54
  Speaker 1: Year before AI can do better than any individual human
• 8:58
  Speaker 1: could do. And then there's a it's a it's a
• 9:00
  Speaker 1: much higher bar to say, well, is this than you know,
• 9:04
  Speaker 1: human intelligence selectively, But if the rate of change continues
• 9:08
  Speaker 1: that that's why I think probably twenty twenty nine or
• 9:10
  Speaker 1: maybe twenty thirty is where digital intelligence will probably exceed
• 9:15
  Speaker 1: all human intelligence combined and then I think it's always
• 9:18
  Speaker 1: helpful to look at these like fundamental ratios, you know,
• 9:21
  Speaker 1: sort of physics first principles approached to looking at things
• 9:24
  Speaker 1: and and.
• 9:27
  Speaker 2: And probabilistic.
• 9:28
  Speaker 1: Yeah, so yeah, it's probablistic, so the Yeah, so if
• 9:31
  Speaker 1: you look at the ratio of digital to biological compute,
• 9:35
  Speaker 1: so like like you know, all of say, all of
• 9:38
  Speaker 1: the higher level cognitive if you sum up the higher
• 9:41
  Speaker 1: level cognitive capacity of humans and then what is the
• 9:45
  Speaker 1: and think of that as compute, then well, and then
• 9:48
  Speaker 1: compare that to what what is the digital compute? And
• 9:53
  Speaker 1: the rate of which this is growing is just buggles
• 9:55
  Speaker 1: the mind. So that's why I think it's you know,
• 9:58
  Speaker 1: I think twenty twenty nine or twenty thirty or thereabouts
• 10:01
  Speaker 1: is is it's not a that's I think a reasonable
• 10:05
  Speaker 1: time frame for where you'd expect the cumulative digital compute
• 10:10
  Speaker 1: to probably exceed the cumulative biological compute of higher level
• 10:13
  Speaker 1: brain functions.
• 10:14
  Speaker 2: And then from from then forever yeah, and still in
• 10:19
  Speaker 2: dispatching and diverging forever from there.
• 10:22
  Speaker 1: Yeah, and then yeah, where do things go from there?
• 10:24
  Speaker 1: I don't know, it probably can continues the we are
• 10:28
  Speaker 1: moving from you say, if you look at the the
• 10:30
  Speaker 1: limiting factors, you know, the what what is the constraint
• 10:34
  Speaker 1: on growth? Like last year it was clearly AI chips
• 10:37
  Speaker 1: were the constraint on growth. Then then this year, the
• 10:42
  Speaker 1: one of the biggest constraints, maybe the biggest constraints on
• 10:45
  Speaker 1: growth are voltage step down transformers, because you know, just
• 10:49
  Speaker 1: just getting the power from like a utility at three
• 10:51
  Speaker 1: hundred killer vaults all the way down to a low
• 10:53
  Speaker 1: one volt for the computer is a massive amount of
• 10:56
  Speaker 1: voltage set down. So it's it's you know, my sort
• 10:59
  Speaker 1: of very niche and perhaps not that funny joke, is
• 11:02
  Speaker 1: that we need transformers for transformers, say you need FULT
• 11:07
  Speaker 1: transformers or AI neural net transformers. That is literally the
• 11:11
  Speaker 1: issue this year. And then if I'm saying like next
• 11:16
  Speaker 1: year and years beyond that, it's actually just it's going
• 11:18
  Speaker 1: to be it's electrical power. And you've got both AI
• 11:23
  Speaker 1: with very big demands for electrical power and the transition
• 11:27
  Speaker 1: to sustainable energy with electric vehicles whatnot also needing electrical power.
• 11:32
  Speaker 1: So it's just a lot of electrical power aids. You know.
• 11:36
  Speaker 2: Elon. One of the one of the things that you
• 11:39
  Speaker 2: said early on when you when you founded Neuralink, which
• 11:42
  Speaker 2: has been amazing. Congratulations on that. And we talked about
• 11:45
  Speaker 2: this with Ray yesterday, talked about Hi Ben with BC.
• 11:48
  Speaker 2: I was I wouldn't put words in your mouth, but
• 11:50
  Speaker 2: I would say it would be more along the ones.
• 11:52
  Speaker 2: If you can't beat them, join them. When it comes to,
• 11:55
  Speaker 2: you know, merging the neo cortex and the cloud, can
• 11:58
  Speaker 2: you I'm looking forward to it. Which is curious what
• 12:00
  Speaker 2: your thoughts are about what's driving that. I meant adding
• 12:04
  Speaker 2: adding that additional computational capacity and sensory capacity to the
• 12:09
  Speaker 2: neo cortex.
• 12:10
  Speaker 1: Yeah, I mean again, this is actually something that in
• 12:13
  Speaker 1: banks in the culture books, which I really recommend everyone
• 12:17
  Speaker 1: read that. In the culture books, there's something called a
• 12:20
  Speaker 1: neural lace, So all the humans have this neural lace
• 12:24
  Speaker 1: that's kind of a very essentially a high band with
• 12:26
  Speaker 1: brain to computer interface. And in the and these are
• 12:30
  Speaker 1: the culture books, that it's it's so good that it
• 12:32
  Speaker 1: actually retains all of your memories and kind of brain states.
• 12:35
  Speaker 1: So even if your physical body dies, you can kind
• 12:37
  Speaker 1: of reincorporated in another physical body and retain you know,
• 12:41
  Speaker 1: pretty much your original memories and and brain state. So
• 12:45
  Speaker 1: now a long way from that, but we only just
• 12:48
  Speaker 1: had our first neural link in the human which is
• 12:50
  Speaker 1: going is going quite well. The first patient is actually
• 12:55
  Speaker 1: able to control their computer just by thinking. Like this
• 12:59
  Speaker 1: first the product we call telepathy, where you can control
• 13:03
  Speaker 1: your computer and phone and through and through your computer
• 13:05
  Speaker 1: and phone almost anything just by thinking. Do you just
• 13:09
  Speaker 1: lie there and think and you can move them at
• 13:11
  Speaker 1: the mouse coaster around the screen and things like you
• 13:15
  Speaker 1: know that we're gonna do. The patient has agreed to
• 13:17
  Speaker 1: do sort of like I think a live demo of
• 13:20
  Speaker 1: just it is quadriplegic where he literally is just controlling
• 13:23
  Speaker 1: the screen. He can like play video games, download software,
• 13:28
  Speaker 1: like really anything you can do with the mouse just
• 13:30
  Speaker 1: by thinking, which is pretty wild.
• 13:32
  Speaker 2: It is, it is pretty wild.
• 13:35
  Speaker 1: Let's turn, I should say, there's a long way to
• 13:38
  Speaker 1: go from that to a whole brain interface. So current
• 13:43
  Speaker 1: new link just has a thousand electrodes. I think ultimately
• 13:46
  Speaker 1: you need something which which has you know, probably one
• 13:49
  Speaker 1: hundred thousand or a million electrodes. No, these are very tiny,
• 13:52
  Speaker 1: very tiny electrodes that they're tiny wires, wastefuler than human hair,
• 13:57
  Speaker 1: and you know, so there's this I just want to
• 13:59
  Speaker 1: upside a long way from from where you're like this
• 14:02
  Speaker 1: today to having a whole brain interface like like the
• 14:04
  Speaker 1: New Release and the Banks novels. But this, this is
• 14:08
  Speaker 1: definitely physically possible, and you know it's sort of kind
• 14:12
  Speaker 1: of like if you can't beat him, join them, you know.
• 14:14
  Speaker 1: So you know, human brain, which has a lot of constraints,
• 14:18
  Speaker 1: it's you know, it's it's yeah, we only have about
• 14:22
  Speaker 1: but maybe ten watts of higher brain function, and we
• 14:25
  Speaker 1: do a lot with our little ten watts. It's not
• 14:27
  Speaker 1: you know, it's progressive. You know that we've voization with
• 14:33
  Speaker 1: such a low power computer. Really it is you know,
• 14:36
  Speaker 1: I sort of think it's like it's not bad for
• 14:39
  Speaker 1: a bunch of monkeys, you.
• 14:39
  Speaker 2: Know, getting some good lass from that. And we've all
• 14:42
  Speaker 2: watched you go from the roadster to the Model three
• 14:45
  Speaker 2: and why and from Falcon one to Starship. So I
• 14:49
  Speaker 2: think going from the first implants to something that's got
• 14:52
  Speaker 2: more capacity, it's just just a matter of if not.
• 14:55
  Speaker 2: When let's talk to.
• 14:57
  Speaker 1: I think ultimately, like I said, you'll have like you'll
• 15:01
  Speaker 1: you'll have kind of a whole brain interface that I
• 15:04
  Speaker 1: guess it is a sort of peraps a form of immortality,
• 15:07
  Speaker 1: and that if it can kind of upload your brain
• 15:13
  Speaker 1: state to you know, if your brain state is essentially stored,
• 15:17
  Speaker 1: you're kind of backed up on a hard drive. I
• 15:18
  Speaker 1: suppose then you know, you can always restore that brain
• 15:21
  Speaker 1: state into a biological body or maybe a robot or something.
• 15:25
  Speaker 1: Then I want to have size again. It's like, you know,
• 15:28
  Speaker 1: many years in the future, but we're not breaking any
• 15:31
  Speaker 1: loads of physics. I can think this is this is
• 15:33
  Speaker 1: probably something that will happen the rate we're building digital superintelligence.
• 15:38
  Speaker 1: It may just be that, you know, we'll have digital
• 15:40
  Speaker 1: super intelligence and it'll just solve the solve the problem
• 15:43
  Speaker 1: for us. But in the meantime, we'll keep progressing with
• 15:46
  Speaker 1: our meat computers and trying to try to do as
• 15:49
  Speaker 1: good as possible.
• 15:51
  Speaker 2: I'm I was going to say, the tools that we
• 15:53
  Speaker 2: have are growing at a super exponential rate that are
• 15:56
  Speaker 2: making our linear projections of the future seem boring in
• 15:59
  Speaker 2: some way. Is one last topic, my friend, which is
• 16:03
  Speaker 2: where you and I first connected in the world of space.
• 16:05
  Speaker 2: Congratulations on Starship three. Amazing, amazing flight, just really spectacular
• 16:11
  Speaker 2: and we all saw a Falcon nine launch from Vandenberg
• 16:14
  Speaker 2: last night, so that was great, and just again thankful
• 16:20
  Speaker 2: for the work you're doing. You know, it's fascinating because
• 16:24
  Speaker 2: I grew up at the late stages of the Apollo
• 16:27
  Speaker 2: and into into the into the Shuttle program, and I
• 16:31
  Speaker 2: can't imagine that any government would be pushing space as
• 16:36
  Speaker 2: rapidly and dramatically as you are, And so thank you
• 16:42
  Speaker 2: for what you're doing there. That's all I can say.
• 16:44
  Speaker 1: Absolutely. Well, I mean, the goal of space sex is
• 16:48
  Speaker 1: is it's just a much bigger goal than any any
• 16:50
  Speaker 1: GOT program, which is to brockets the spacecraft that are
• 16:54
  Speaker 1: capable of making life multiplanetary. So you know, I mean
• 16:58
  Speaker 1: step one is actually having that as a goal. If
• 17:01
  Speaker 1: you don't have that as a goal, you're definitely not
• 17:02
  Speaker 1: going to achieve it. If you have it as a goal,
• 17:05
  Speaker 1: now at least you have a chance of achieving it.
• 17:07
  Speaker 1: And the thing about Starship is it is it is
• 17:11
  Speaker 1: the first rocket where making life multiplanetary and you have
• 17:15
  Speaker 1: puilding a self sustaining city on Mars is at least possible.
• 17:19
  Speaker 1: It's still obviously an immense amount of work, but it
• 17:21
  Speaker 1: is the first rocket where that is success in making
• 17:26
  Speaker 1: light multiplanetaries at least one of the possible outcomes.
• 17:29
  Speaker 2: Yeah, I'm wondering if you're willing to venture a guest
• 17:31
  Speaker 2: on when you'll be on the moon.
• 17:32
  Speaker 1: I think pretty soon. I'd be surprised if it's longer
• 17:36
  Speaker 1: than about three years to be landing starships on the Moon.
• 17:39
  Speaker 1: And because the progression of Starship is very rapid, you know,
• 17:43
  Speaker 1: we're hoping to do at least another maybe five or
• 17:46
  Speaker 1: six flights this year, and with each successive flight making
• 17:49
  Speaker 1: significant improvements. So I think we've got to do a
• 17:53
  Speaker 1: decent shot of achieving full reusability of both stages of
• 17:58
  Speaker 1: booster around the ship this year, and if not this year,
• 18:01
  Speaker 1: I think, you know, knock on wood, It's like, I
• 18:03
  Speaker 1: think it's a very high probability of achieving full reusability
• 18:07
  Speaker 1: next year, which really is the fundamental breakthrough needed to
• 18:10
  Speaker 1: make life multiplanetary. Yeah, for for those that that that
• 18:14
  Speaker 1: that don't don't know the rocket industry, that that well
• 18:18
  Speaker 1: that they may not be aware that that that this
• 18:20
  Speaker 1: is really the holy grail of rocketry is is full,
• 18:22
  Speaker 1: full and rapid reusability. Because at that point you're you're
• 18:26
  Speaker 1: really just constrained by your propellant costs and Starship, you know,
• 18:31
  Speaker 1: almost eighty percent of the propellant is liquid, oxygen which
• 18:34
  Speaker 1: is very low cost, and then the fuel has met
• 18:37
  Speaker 1: the sort of little over twenty percent of fuel which
• 18:39
  Speaker 1: is sort of methane, also the lowest cost fuel. So
• 18:41
  Speaker 1: if you have full of rapid reusability, then your your
• 18:45
  Speaker 1: actual cost per flight of starship. Even though it's it'll
• 18:48
  Speaker 1: be capable of we think ultimately two hundred tons to
• 18:51
  Speaker 1: orbit will be maybe all the two.
• 18:55
  Speaker 2: It's too