Latest / Elon Musk Podcast / Elon Musk: Will AI & Robots Make Your Job Obsolete? How to Stay Valuable in the Next 20 Years

Transcript

Elon Musk: Will AI & Robots Make Your Job Obsolete? How to Stay Valuable in the Next 20 Years · Elon Musk Podcast

• 0:00
  Thanks to our amazing community members like you, we've reached
• 0:04
  the top 15 of Spotify's video podcast, the Top 10 audio
• 0:08
  podcasts on both Apple and Spotify for the tech category,
• 0:12
  so you all make this possible. If you want to support us more,
• 0:16
  check out our Patreon that's patreon.com/stage 0 News so we
• 0:20
  can keep this free and open for you to enjoy.
• 0:26
  Something interesting happened the other day.
• 0:28
  I was looking through our stats on Spotify and Apple Podcasts
• 0:31
  and I noticed that about 55% of you are not subscribed to the
• 0:37
  show. That means 45% of you are
• 0:39
  subscribed and I really do appreciate your support.
• 0:42
  Now the other 55% of you are awesome, but I'm going to ask
• 0:47
  you for a favor. Could you please hit the
• 0:48
  subscribe button? It'll take you one second, but I
• 0:51
  promise you 10 years of this podcast for free, no pay walls.
• 0:56
  I'm not going to charge you anything ever, but I'm going to
• 0:58
  give you 10 years of this show for free.
• 1:00
  I've already been doing it for five years and I plan on doing
• 1:03
  it for 10 more. And the only way that we can
• 1:05
  continue doing this is with your support.
• 1:08
  So one second of your time to hit the subscribe button right
• 1:11
  now would help the show tremendously.
• 1:14
  Thank you so much. Elon Musk says AI and robots
• 1:19
  will make most job skills obsolete within 20 years.
• 1:23
  And he still wants his kids to go to college, though that is
• 1:26
  the core fact, and it sets the stakes for anyone planning
• 1:29
  school or a career right now. He pictures a world where
• 1:32
  machines handle nearly all the work society needs, from routine
• 1:36
  tasks to a lot of knowledge work.
• 1:38
  He also says college can still be useful for social growth and
• 1:41
  broad learning. If work could be optional.
• 1:45
  What does college actually buy you?
• 1:48
  After this breaks, I'm going to unpack this in plain terms for
• 1:51
  you, and I'll help you translate this into moves you can make
• 1:56
  right now. And then we're going to talk
• 1:58
  about a short plan for students, parents, and early career folks
• 2:01
  who want to stay valuable without chasing fads.
• 2:04
  So stick around and we'll get right into it.
• 2:09
  So here it is. Must lays out a near term future
• 2:12
  where AI and robotics take over most tasks people do for pay.
• 2:16
  That pushes us to separate training for tasks from learning
• 2:20
  for life. We're going to frame this as
• 2:23
  college as a place that can still pay off if you use it for
• 2:26
  range judgement and a real network, not just a line on a
• 2:30
  resume. Let's start out with the
• 2:32
  prediction Elon Musk puts the timeline in under 20 years.
• 2:36
  That sits inside one education and career arc for Gen.
• 2:39
  Z and Gen. Alpha.
• 2:40
  A middle schooler today could graduate into an economy where
• 2:43
  entry level task work is thin and supervision of AI systems is
• 2:47
  common. Now, if you accept that window,
• 2:50
  which changes how you plan major certificates and early jobs, now
• 2:56
  you stop trying to be the model that wrote production, and you
• 3:01
  Start learning to direct systems, test outputs, and make
• 3:04
  decisions that carry risk. The AI and the robots will do
• 3:09
  the hard work now, translate optional work into the day
• 3:14
  today. If machines produce goods and
• 3:16
  services cheaply, paid roles shift toward the humane parts of
• 3:19
  value. People define goals, set
• 3:21
  standards, review outputs, and accept accountability.
• 3:25
  It's different from doing every step yourself.
• 3:27
  You need judgement, not only speed.
• 3:30
  You need to decide what good it looks like, not only generate a
• 3:33
  first draft. You need to explain choices to
• 3:36
  other people who also have tools.
• 3:38
  That is a lane where humans still matter now.
• 3:42
  So why would Elon Musk still back college if tools get that
• 3:46
  strong now? There's two reasons that pass
• 3:49
  that test. First, social growth, which
• 3:51
  means a dense network appears and mentors you can practice
• 3:54
  with. You sit in rooms with people who
• 3:56
  challenge you, and you learn to work in groups and under
• 3:59
  pressure. Second, broad learning, which
• 4:02
  means you see many fields before you settle on a track.
• 4:05
  Range helps you connect ideas when tools change.
• 4:08
  And if narrow skills expire fast, range is a hedge that
• 4:12
  keeps paying. Now you might hear that and
• 4:16
  think learning is pointless if AI can answer everything.
• 4:21
  Now that is not what the educators argue here.
• 4:24
  The problem is not learning. The problem is shallow
• 4:27
  assignments. If a course rewards formulaic
• 4:30
  output, a model will do it in seconds and the student will
• 4:34
  learn nothing. The fix is to raise the bar.
• 4:37
  You force analysis. You ask for evidence and a real
• 4:40
  choice explained in clear language.
• 4:42
  You make the student own the conclusion and in that setup and
• 4:46
  it becomes a starting point, not a substitute.
• 4:50
  Now let's make this concrete for students right now.
• 4:53
  Pick a field you like. So if you break that job into
• 4:56
  tasks, mark the tasks a model can already do and the ones that
• 5:02
  still need judgement. Put your time on the judgement
• 5:05
  tasks. Learn to design prompts that
• 5:08
  collect the structure you need, and then you build a workflow
• 5:11
  that combines tools and checks. You audit the outputs with a
• 5:14
  simple test list. Ask what assumption the model
• 5:17
  made, ask what context it missed, ask how you would know
• 5:21
  if the answer hurt someone or wasted money.
• 5:24
  Now that motion turns raw AI speed into reliable human
• 5:29
  results. And we think about this.
• 5:31
  Proof beats postures, so change how you show your work too.
• 5:36
  You can build a portfolio with full artifacts.
• 5:39
  Include the prompt change you used, the mistakes you found,
• 5:42
  the fixes you tried, and the call you made at the very end at
• 5:45
  a one page reflection that says what you learned and what you
• 5:49
  would change next time. That packet tells a hiring
• 5:52
  manager that you do not just press a button.
• 5:55
  You don't just prompt something. You actually have a system.
• 5:58
  You plan, you test, you decide, and you can explain your
• 6:01
  decision. That is the job.
• 6:03
  When tools get good, when AI starts taking over the low jobs,
• 6:10
  you need to be there to make those decisions.
• 6:12
  And parents always ask what to emphasize at home.
• 6:15
  If you're a parent, it's important that you talk to your
• 6:19
  kids about this stuff. AI is taking over everything.
• 6:22
  They use it in school now, they have it on their phones, They
• 6:25
  have chat TBT in their pocket. So you have to start with range.
• 6:29
  Read outside your bubble. Mixed science, history,
• 6:32
  economics, and philosophy into your reading.
• 6:35
  Range gives you patterns to compare model outputs against.
• 6:39
  Encourage public work to A paper that one teacher reads builds
• 6:44
  one set of muscles. A project that 200 strangers
• 6:48
  read builds another. Public work forces clarity,
• 6:52
  honest error handling. You got to know your mistakes
• 6:55
  and you got to own them. And a basic respect for the
• 6:58
  audience. Those are the muscles that carry
• 7:00
  into any field of learning, any field of work, and any sort of
• 7:05
  prompting. It'll mix your mind in a way
• 7:08
  that will allow you to prompt things that nobody else can.
• 7:11
  Be unique, be individual, and make sure that your chain is
• 7:15
  good now. Early career folks want leverage
• 7:19
  fast too. This.
• 7:21
  There's some logic to this. Treat AI fluency as a baseline.
• 7:27
  Then you ask for the ownership. You can volunteer to define
• 7:30
  success metrics on a project. Offer to write the post mortem.
• 7:34
  When something fails, you can propose the next test.
• 7:38
  Learn enough of the math and data to spot what a result looks
• 7:40
  wrong. Learn to say what you will do
• 7:44
  about it. How you got to fix this.
• 7:46
  Ownership is where compensation and stability live.
• 7:50
  If you can tell people that you're wrong and you can own it,
• 7:54
  people will respect you. Because someone needs to be
• 7:57
  responsible when the decision matters.
• 7:59
  And if that person is going to be you, even if you are wrong,
• 8:02
  that's OK. Now let's talk about
• 8:05
  entrepreneur in this center ship.
• 8:08
  If tools make production cheap, the constraint becomes taste,
• 8:12
  distribution, and trust. A small team can ship a product
• 8:16
  if it can reach an audience and keep the promises of the
• 8:19
  original founders. You want to talk to your users?
• 8:23
  I'm going to read their feedback and start shaping your features
• 8:28
  that fit their use. Now this is where college is
• 8:31
  helpful. It gives you a lab where peers
• 8:35
  and mentors are. A non college path can work if
• 8:39
  you put yourself in communities where people build in public.
• 8:42
  You trade your notes, you put each other to improve and be
• 8:46
  innovative. There's also a civic piece that
• 8:49
  we should say we talked about. If paid work declines for many
• 8:53
  people, communities still need leaders.
• 8:55
  Institutions need people who can set norms, resolve conflicts,
• 9:00
  and steward those resources that everybody's going to be using.
• 9:04
  Families and schools can prepare students by giving them real
• 9:07
  responsibilities early on. Let them host an event.
• 9:12
  Let them run a small budget. Let them learn to handle
• 9:14
  mistakes without hiding them. Own your mistakes, people.
• 9:18
  People will respect you more, I promise you.
• 9:20
  I've been in business for the last 30 years.
• 9:24
  I've been doing it my whole life.
• 9:26
  I've been a contractor for the last 10 or 15 years and I've
• 9:33
  been outsourcing other jobs to other people.
• 9:36
  I've been managing people. I've run 3 or 4 different
• 9:40
  companies, sold off 3 or 4 companies as well.
• 9:43
  I've been Aqua hired. I've got a big payout at a tech
• 9:48
  firm. So this really helps.
• 9:52
  If you can own your mistakes, you're going to be so much more
• 9:55
  trusted than somebody who doesn't make mistakes or
• 9:59
  somebody who just doesn't own up to it.
• 10:01
  People want real people involved with these.
• 10:04
  Don't hide your mistakes. And the big risk now is false
• 10:08
  certainty. You don't know what's going to
• 10:09
  be happening, so don't even try this.
• 10:11
  Some people will assume AI will never do their job, then kablam.
• 10:16
  You get blindsided by a tool that does it faster and cheaper.
• 10:19
  I was a victim of that, right? I knew that AI was coming for
• 10:25
  coders. So if you're a high level coder,
• 10:28
  if you're a senior level coder right now you're doing OK, But
• 10:33
  in the next 5 or 10 years, you're going to be a prompter.
• 10:37
  You're not going to be coding by hand anymore because tools like
• 10:41
  Replit exist. Tools like Lovable exist and you
• 10:46
  can literally create a whole chain of prompts that will build
• 10:52
  you an enterprise level application in a weekend.
• 10:57
  I've done it numerous times. I've built 5 or 6 applications
• 11:02
  in a weekend and they were all pretty decent things that I
• 11:06
  couldn't do in the back end. Replica build, Riplet REPLIT.
• 11:11
  If you're a coder, you've got to know about it.
• 11:14
  It's a tool that you're going to be able to use in the future and
• 11:19
  you're going to be using it after this episode because you
• 11:22
  need to, absolutely need to, because if you need something
• 11:26
  built, it'll build it for you. No joke.
• 11:29
  You can take all the time you want, build it by hand, that's
• 11:31
  fine, that's understandable. But if something can build it
• 11:35
  for you and it can do it pretty damn good, I would say 90% of
• 11:42
  the time it gets it. It's a great product.
• 11:46
  And those other 10%, you got to fix some stuff and re prompt and
• 11:49
  things. But it's the same thing with
• 11:51
  code. There's errors, there's bugs and
• 11:54
  you'll find those and you'll have to recode it yourself and
• 11:56
  that's for now until these apps get even better.
• 12:01
  Give it a year. You're going to see some things
• 12:03
  that you won't even believe. You'll be able to write a
• 12:05
  prompt. Hey, I want to make it XYZ and
• 12:08
  it'll spit it out for you in a seconds.
• 12:12
  Now you have to prompt re prompt.
• 12:15
  It asks for certain things like authentication from other apps
• 12:19
  like Stripe and etcetera. So that is one of the things
• 12:24
  that blindsided me. I knew it was coming, but it
• 12:27
  came so fast. I'm still OK, I can still code
• 12:31
  and I can still do all this stuff, but it the market has
• 12:34
  just dropped out of the early level developers.
• 12:38
  You have to be either a senior level developer or somebody who
• 12:42
  understands prompting so well that you can build applications
• 12:45
  on the fly. So the uncertainty could just
• 12:49
  destroy you and other people the give up on learning because a
• 12:55
  model already exists for it. Like why would I learn how to
• 12:57
  code? This thing codes for me.
• 13:01
  But that's a mistake. The point is not to defend a
• 13:04
  current task list that you have right now.
• 13:06
  The point is to deliver value that people can trust.
• 13:10
  I've been doing this for 1520 years, I've been coding 1520
• 13:15
  years, and I've also been developing content for the last
• 13:20
  30 years from network television to low budget skateboarding
• 13:26
  films to corporate gigs, things, the whole spectrum.
• 13:32
  People can trust me when it comes to developing content for
• 13:36
  them, editing video, shooting with a camera, things like that.
• 13:40
  So people can trust me in my industry.
• 13:43
  If you want to know more about what I do, go to willwalden.com,
• 13:46
  WILWAL, don.com and you'll learn all about the things that
• 13:51
  ioffer. Now if you build for judgement
• 13:54
  though range and ownership, you can move when the tools move and
• 13:58
  still hold your ground because these things are ever evolving.
• 14:01
  Once the tools get better, they're just going to get wild
• 14:05
  because the tools will build the tools to the machine that builds
• 14:08
  the machine. And once these AI tools get so
• 14:12
  good they can build themselves even better, then you really
• 14:17
  will get blindsided. So Elon predicts that AI and
• 14:22
  robotics will make most job skills obsolete within 20 years.
• 14:26
  Those are the skills that are available right now.
• 14:28
  Robotics will happen and it will happen really fast too.
• 14:32
  In the next 5 years, I predict that there will be a big shift
• 14:39
  in robotics in the home. Things like a dedicated sort of
• 14:46
  like robot dishwasher that only does 1 task, right?
• 14:50
  You might not need the whole robot.
• 14:51
  You might not need an optimist. Elon says they're going to be
• 14:54
  $20,000. I highly doubt they're going to
• 14:57
  be $20,000 by the time it gets to a point where the robot is
• 15:03
  good enough to do all the tests that Elon wants it to do.
• 15:07
  Do your dishes, hold your clothes, do your laundry,
• 15:11
  etcetera. It's going to cost a lot more
• 15:13
  money than 20,000. So maybe there will be smaller
• 15:16
  robots. I mean, we already have robots
• 15:19
  that vacuum our floors and they do it on a schedule.
• 15:24
  They do it every every morning at 8:30 AM or something.
• 15:31
  And that's all you really need. You don't need a robot to push
• 15:34
  your vacuum. You don't need that.
• 15:36
  You have a, you already have a robot.
• 15:38
  That's so vacuum robot. Now if you could find a robot
• 15:41
  that does your dishes for you and does your laundry, that
• 15:44
  would be great. And there will be systems set up
• 15:46
  for that I'm sure. Not just an all around general
• 15:49
  purpose robot like Optimist. I'm sure there are other robots
• 15:52
  that are going to take all the heartache away from you.
• 15:55
  All that stuff that you don't want to do.
• 15:58
  I did. I washed the dishes today.
• 16:01
  I did the laundry, I did the floors, I made the bed, I did my
• 16:06
  bedding, I cleaned up, tidied up.
• 16:10
  And if you have little robots that run around, little R2D twos
• 16:15
  that run around, that might be the the source or you have one
• 16:18
  robot like an Optimist. But if you can afford 20 to
• 16:22
  $50,000 or can you afford $2000, that's where it come what it
• 16:27
  comes down to a robot vacuum right now under $1000.
• 16:32
  That saves you so much time. Never have to vacuum and that
• 16:37
  time can be spent with your friends, with your family
• 16:39
  outside, doing something, making your next big project.
• 16:43
  Anyway, if you're a business to business owner and you need help
• 16:49
  getting your product on YouTube or getting your message on
• 16:53
  YouTube, I can help you out. Go to willwalden.com now.
• 16:57
  I want you to succeed with your AI plan.
• 17:01
  Use AI for speed. Keep humans for the judgement.
• 17:05
  You build the range of that person, build your mind and
• 17:08
  prove you own those outcomes with your products, with your
• 17:14
  services, things that you've done for a long time, things
• 17:16
  that you can prove, and then you're going to be OK.
• 17:22
  Support. If you could take a second and
• 17:23
  hit the subscribe or the follow button on whatever podcast
• 17:26
  platform that you're listening on right now, I greatly
• 17:29
  appreciate it. It helps out the show
• 17:31
  tremendously and you'll never miss an episode.
• 17:34
  And each episode is about 10 minutes or less to get you
• 17:37
  caught up quickly. And please, if you want to
• 17:40
  support the show even more, go to patreon.com/stage Zero.
• 17:46
  And please take care of yourselves and each other, and
• 17:48
  I'll see you tomorrow.