Latest / Elon Musk Podcast / Elon Musk talks about the future of Twitter/X

Transcript

Elon Musk talks about the future of Twitter/X · Elon Musk Podcast

• 0:01
  Hey everybody, welcome back to the Elon Musk Podcast.
• 0:05
  This is a show where we discuss the critical Crossroads, The
• 0:08
  Shape, SpaceX, Tesla X, The Boring Company, and Neuralink,
• 0:13
  and I'm your host, Will Walden. If you want uninterrupted
• 0:17
  episodes of the Elon Musk podcast, please go to
• 0:20
  clubelon.supercast.com to find out how there's a link in the
• 0:26
  show notes. If you if you empower people as
• 0:29
  censors, then well, there's going to be some amount of bias
• 0:33
  they have and then whoever appoints the censors is
• 0:36
  effectively in control of information.
• 0:39
  So then the idea behind community notes is well how do
• 0:42
  we have a consensus driven, I mean so it's not really
• 0:47
  censoring it but consensus driven approach to truth.
• 0:50
  How do we or how do we, how do we make things the least amount
• 0:54
  untrue? Like you can say like you can't
• 0:57
  pass perhaps get to pure truth but you can aspire to be more
• 1:01
  truthful. So the the the thing about
• 1:04
  community notes is it doesn't actually delete anything.
• 1:07
  It simply adds context. Now that context could be this
• 1:09
  thing is untrue for the following reasons and but but
• 1:15
  importantly with community notes everything is open source
• 1:18
  actually. So you can see the the software,
• 1:22
  every line of the software, you can see all of the data that
• 1:24
  went into a community node and you and you can independently
• 1:27
  create that community node. So if you've got if if you see
• 1:30
  manipulation of the data, you can actually highlight that and
• 1:33
  say well this, this, this, there appears to be some gaming of the
• 1:36
  system and you can suggest improvements.
• 1:40
  So it's it's it's maximum transparency, which is I think
• 1:44
  combined with the kind of wisdom of the crowds and transparency
• 1:47
  to get to a better answer. And really, one of the key
• 1:50
  elements of community notes is that in order for a note to be
• 1:52
  shown, people who have historically disagreed must
• 1:55
  agree. And there is a bit of AI usage
• 1:59
  here. So this will populate A
• 2:01
  parameter space around each contributor to community notes
• 2:07
  and then a parameter space. So, so everyone's got basically
• 2:11
  these these vectors associated with them which so it's it's not
• 2:14
  as simple as as right or left it's saying it's it's more it's
• 2:18
  several 100 vectors that that because things are more
• 2:21
  complicated than something right or left and and then we'll we'll
• 2:25
  do sort of inverse correlation say like OK these these people
• 2:31
  generally disagree but they agree about this note OK so then
• 2:34
  that so then that that that gives the note credibility.
• 2:37
  OK yeah that's that's the that's the core of it and it's working
• 2:40
  quite well. I get to see a note actually be
• 2:45
  be present for more than a few hours that that is incorrect.
• 2:48
  So the batting average is extremely good and and when I
• 2:50
  ask people say oh they're worried about community notes
• 2:53
  sort of being disinformation like send me one and then they
• 2:56
  can't so. So I think it's I think it's
• 3:00
  quite good. I mean the general aspiration is
• 3:02
  with with the X platform is to inform and entertain the public
• 3:07
  and to be as accurate as possible and as truthful as
• 3:09
  possible even if someone doesn't like the truth.
• 3:12
  You know people don't always like the truth not always but
• 3:19
  but that's that's that's the aspiration and I think if if we
• 3:22
  are if we stay true to the truth then I think we'll find that
• 3:25
  people use use the system to learn what is going on and to to
• 3:31
  I think actually truth pays. So I I think it'll be what what
• 3:37
  I mean assuming you don't want to engage in self delusion then
• 3:42
  then I think it's it's the smart move excellent right that there
• 3:48
  are cultural elements where you you you know the the culture
• 3:53
  should celebrate creating new companies and and and there
• 3:59
  should be a bias towards supporting small small companies
• 4:04
  because they're they're the ones that need nurturing.
• 4:06
  The larger companies really don't need nurturing.
• 4:10
  So you know just you can think of sort of like a garden if it's
• 4:13
  a little sprout that needs needs nurturing if it's a mighty oak
• 4:16
  it doesn't need quite as much. So I think that that that is a
• 4:20
  mindset change that is important but but I should I should
• 4:24
  mention that London is you know London and San Francisco or the
• 4:30
  Bay Area are really the two Centers for for AI.
• 4:33
  So that so London is actually doing doing very well on that
• 4:36
  front that the two I say the two leading locations on Earth, you
• 4:41
  know San Francisco's probably head of London but London's
• 4:44
  really very strong or London area, Greater London Home
• 4:48
  Counties I guess keep going keep going.
• 4:54
  So I'm just saying objectively this is the case and but you do
• 5:00
  need that. You need, you need the
• 5:02
  infrastructure. You need, you need landlords who
• 5:04
  are willing to rent to new companies.
• 5:06
  You you need law firms and accountants that are willing to
• 5:09
  support new companies. And it's generally a might it is
• 5:11
  a mindset change and I think some of that is happening.
• 5:15
  But I think really it's just culturally people need to decide
• 5:18
  this is this is a good thing. Yeah.
• 5:20
  Exceed with your first start up. It shouldn't be a sort of a
• 5:23
  catastrophic career ending thing.
• 5:25
  It should be you know, well good.
• 5:27
  I think generally should like it's really like well you know
• 5:29
  you gave it a good shot you know and and and now try again yeah
• 5:33
  and it's so one thing I was going to mention is like
• 5:37
  obviously creating a company is sort of a high risk high reward
• 5:39
  situation but I don't know quite what the how it works in in the
• 5:45
  UKI think it's probably better than than Continental Europe but
• 5:49
  the the stock options are very difficult in most parts of
• 5:53
  Europe. I'm not sure where how does new
• 5:55
  game but if somebody's basically going to risk their their sort
• 5:58
  of life savings and with and the vast majority of startups fail.
• 6:03
  So I mean you hear about the startups that succeed, but most
• 6:06
  companies are, most startups consist of, you know, a massive
• 6:10
  amount of of of work followed by failure.
• 6:15
  That's actually most, most companies.
• 6:17
  And so it's a high risk, high reward.
• 6:18
  And so the high reward part does need to be there for it to make
• 6:22
  sense. Yeah, exactly.
• 6:24
  The most immediate thing is just being able to ask, like having a
• 6:28
  very smart friend that you can ask anything, you know, ask how
• 6:33
  to make something, how to solve any problem and it'll tell you
• 6:38
  so. And obviously companies are
• 6:41
  going to adopt this. So I think you'll have much
• 6:44
  better customer service. I guess essentially that'll
• 6:46
  probably be the first thing you'll notice.
• 6:49
  And and then we talked about education, so having a tutor.
• 6:57
  So if you're trying to understand a subject like having
• 7:01
  a phenomenal tutor on any subject it's that that's really
• 7:06
  pretty much there already almost.
• 7:07
  I mean we need to obviously AI need to stop hallucinating
• 7:10
  before you know it can't give you.
• 7:13
  I mean we still have a little bit of a problem where it can
• 7:16
  give you an answer that's confidently wrong with great
• 7:19
  grammar and you know in a bullet points and everything in
• 7:23
  citations it was not real. So that's me.
• 7:27
  OK, we need to make sure it's it's not.
• 7:29
  It's not. It's not to giving you
• 7:30
  confidently wrong tutor answers, but that's going to happen
• 7:34
  pretty quickly where it is actually correct.
• 7:39
  So, well, really anything that can be actuated by a computer is
• 7:45
  effectively a robot. So you can think of, frankly,
• 7:49
  Tesla cars are robots on wheels. Anything that's connected to the
• 7:54
  Internet is effectively an endpoint actuator for artificial
• 7:58
  intelligence. So you've got Boston Dynamics.
• 8:06
  Obviously they've been making impressive robots for a while.
• 8:09
  I think they're at this point mostly owned by Hyundai.
• 8:12
  So I I guess Hyundai is probably going to make robots of that are
• 8:16
  humanoid and and and some rather interesting shapes that I wasn't
• 8:20
  just being like the one that looks like a has wheels and
• 8:23
  looks sort of like a kangaroo on wheels.
• 8:25
  I'm not sure what that is but looks a little demanded frankly
• 8:31
  but but there's going to be all sorts of all sorts of robots.
• 8:37
  You've got the company Dyson in in the UK which I think there's
• 8:40
  some pretty impressive things. I I think the UK will not be
• 8:46
  behind actually on on that front.
• 8:48
  UK also has ARM which is really the the best one of the best,
• 8:54
  perhaps the best in in chip design in the world.
• 9:01
  Tesla uses a lot of a lot of ARM technology, almost everyone does
• 9:05
  actually. So I think UK is in a in a
• 9:08
  strong position. Germany also makes a lot of
• 9:11
  robots, national robots, I mean I think generally countries that
• 9:15
  make robots of any kind, even if they seem somewhat conventional
• 9:21
  will be, will be fine. I do think there is a, there is
• 9:26
  a a safety concern especially with humanoid robots because you
• 9:31
  know at least the car can't chase you into this building.
• 9:34
  Not very easily you know or chase you up a tree or you know
• 9:39
  you can sort of run up a flight of stairs and get away from a
• 9:41
  Tesla. I think it's a Stephen King
• 9:45
  movie about that if your car gets possessed so but if you
• 9:51
  have a humanoid robot it can, it can basically chase you
• 9:54
  anywhere. So I think we should have some
• 9:57
  kind of hardwired local cut off that you can't update from the
• 10:03
  Internet. So anything that can be software
• 10:06
  updated from the Internet obviously can be overridden.
• 10:08
  But if you have a local sort of off switch where you perhaps say
• 10:12
  a keyword or something, and then that puts the robot into a safe
• 10:15
  state, some kind of localized safe state ability, an off
• 10:23
  switch, you know, or you don't have to get too close to the
• 10:27
  robot. I know.
• 10:27
  So if we've got millions of these things going all over the
• 10:30
  place. Hey, thank you so much for
• 10:33
  listening today. I really do appreciate your
• 10:35
  support. If you could take a second and
• 10:37
  hit the subscribe or the follow button on whatever podcast
• 10:40
  platform that you're listening on right now, I'd greatly
• 10:43
  appreciate it. It helps out the show
• 10:44
  tremendously and you'll never miss an episode and each episode
• 10:48
  is about 10 minutes or less to get you caught up quickly.
• 10:51
  And please, if you want to support the show even more, go
• 10:55
  to Atreoncom, Stage Zero, and please take care of yourselves
• 11:00
  and each other and I'll see you tomorrow.