Ex Tesla President Shares His AI Predictions for 2026

0:00 What do you think now with AI?

0:01 So, uh, I feel like you've been doing

0:03 this for a long time and just when you're probably like,

0:05 I'm a veteran of this game, the entire board game changes.

0:08 Somebody flips the table and now there's a whole new game to be played.

0:12 Where's your head at with AI?

0:13 How do you think about this, man?

0:14 I'm excited.

0:15 I all the way back in college,

0:17 I worked my way through college writing trading algorithms at uh

0:21 at the world's largest options and futures trading uh platform in Chicago.

0:26 I went I went to college in Chicago area and worked my afternoons coding and way

0:33 back then we were using neural the very beginning neural nets to try to get

0:37 an edge in the market and try to and try to robo trade uh

0:41 in a way that people hadn't been able to in terms of the accuracy and the speed.

0:46 So all these years later now that this has emerged I'm

0:49 like completely psyched because we were using what looked like like

0:53 rudimentary computers uh to do this stuff.

0:55 Now you got real compute power.

0:57 Caveman AI.

0:59 Totally.

0:59 And now that you see this unleashed, I'm like, "Ah,

1:01 this is what we were trying to trying to do three decades ago." So, um,

1:06 so I'm super excited by it

1:08 largely because every technical revolution and breakthrough

1:12 like this that has happened

1:14 in history creates enormous opportunities for entrepreneurs.

1:19 I cannot name a technical revolution that's happened

1:22 that's resulted in less GDP and less jobs.

1:25 can't like it just doesn't work that way.

1:28 Um, and there's a lot of hand ringing though at the beginning of every

1:31 technological revolution because humans are really

1:34 good at seeing the first order effect

1:35 which is the job destruction but they're not good at seeing the second order

1:39 effect which is the job creation that happens by entrepreneurs on the back end.

1:43 And I think we're like right in the middle of this.

1:45 So I we're we're applying AI at the core of each of our companies.

1:50 But I think more importantly,

1:52 I think there's going to be just like we're keeping our eyes

1:54 peeled for big big opportunities that are being created by this.

1:58 So I give you an example,

1:59 and this isn't this example would date me, but when I was growing up,

2:05 uh to make a longdistance phone call,

2:08 you had to dial zero and talk to a human being.

2:11 And that was usually a lady in your town uh who would

2:14 literally plug the local phone network on one end of a cord

2:18 and then plug that cord into the national phone network

2:20 and that would create a connection for you to do a long-distance call.

2:23 In other words, humans were switches.

2:25 Uh and um sometime in the late '60s, early '7s,

2:30 Bell Labs invented an electronic switch that would do this.

2:34 And by the end of the 70s,

2:37 there was a lot of hand ringing that there were 800,000 people,

2:41 pillars of our communities that were employed as operators.

2:44 And 100% of those jobs were going away.

2:48 And sure enough, by the late '7s, early 80s, 100% of those jobs had gone away.

2:52 They were those humans were replaced by electronic switches.

2:56 But what people couldn't see while they were doing the hand

2:59 ringing was what was going to happen on the other side.

3:02 And what happened on the other side was now that a long-distance call is free,

3:07 doesn't require human labor, just requires electrons,

3:10 now you could offer toll-free calls.

3:13 And so some entrepreneurs said,

3:16 "I'm going to I'm going to use these toll-free calls." There was

3:18 a single area code 1800 uh and I'm going to create businesses.

3:22 1-800 flowers, 1800 junk, 1800 insurance.

3:25 We want to enter this that.

3:27 And what those 1-800 numbers required was now centralized answering of phone

3:32 calls cuz now people were calling like crazy cuz it was free.

3:35 And so this industry got created called centers and support centers.

3:40 And a few years into the 1980s,

3:43 there were now millions of people employed in call centers.

3:47 Hundreds of software firms that were created.

3:50 And my first startup was a software firm in the '9s

3:53 serving call centers uh because the technology was still

3:57 nent when people were just deploying electronic switches and people

4:02 were like lamenting the end of these jobs as operators.

4:05 Nobody could see on the other side

4:07 that these entrepreneurs were going to create toll-free dialing.

4:10 They were going to create businesses on top of toll-free dialing.

4:12 And on top of that, there were going to be this whole layer in our uh

4:16 in our economy called call centers that millions

4:18 of people are going to be employed in.

4:20 We just had creativity to see the other side.

4:22 And so now I tell our teams like we cannot see through the other side,

4:27 but we can get a hint of where

4:28 this stuff is going to go if you think creatively.

4:31 Uh and and so that's the thing I'm most excited about.

4:34 Like we're going to have entrepreneurial opportunities like crazy over the next

4:39 few years based on what this technology is going to unleash.

4:42 I love that story.

4:43 So, keep going.

4:43 So, let's say uh you know, if we can think creatively,

4:45 we could start to get a glimpse of what maybe is possible.

4:49 What do you see as the kind of first generation businesses

4:51 that you're excited about and what the second order effect might be?

4:54 Well, now we've got intelligence that is so rapid and non-latent,

4:58 so you can solve problems way faster that are really complex.

5:02 So, one of the one of the businesses that we've got that I'm most excited about

5:06 is we pulled a team out of Tesla

5:08 that built the supply chain automation platform at Tesla,

5:12 which is super sophisticated and way

5:13 beyond anything anybody had in the industry.

5:16 And they built this with ML uh in uh in like 2017ish.

5:22 And we said, how would you like to build it today in AI?

5:25 And what AI allows that team to do

5:27 is solve a super complex problem really quickly,

5:30 but also uh use Aentic AI to understand

5:34 the workflows of their clients super rapidly.

5:38 So they just went into one

5:39 of the biggest grocery delivery platforms in the country.

5:42 I don't think I can say which one yet publicly,

5:45 but it's it's one of their newer customers.

5:48 and their agents could go in and understand the work rules

5:51 of that ginormous platform within hours and then design a system

5:57 and a workflow within hours that um uh that you really

6:02 have to have two special things to be able to do.

6:04 You have to be expert in that particular thing which

6:09 is supply chain optimization to know which rules are good,

6:12 which rules are bad, what to tell your agents to do etc.

6:15 And then you have to be able to judge, okay,

6:17 what these agents are bringing back is correct

6:18 or it needs to be tweaked or redirected or whatever.

6:21 So you definitely have these humans who have got AI as exoskeleton like

6:26 super strength that are now building a business that's growing like a rapid

6:30 fire because it's got this incredible commute

6:33 compute to actually fuel uh the business

6:37 and deliver value back to the customer way faster than you could otherwise.

6:41 like they what they're competing against is standard ERP systems that take 9

6:45 to 12 months to implement and they can implement in a period of days

6:49 and deliver that value back and um that's the kind of business like I

6:53 think we're going to see more and more

6:55 and more of as this capabilities unleashed.

6:58 Yeah, that is that is the weird thing

7:00 about this that the software essentially works like labor, right?

7:05 So what you're describing is like Yeah, exactly.

7:07 essentially a consultant.

7:08 It comes in and it learns your workflows.

7:09 It documents it.

7:10 figures out how things are currently working and then

7:12 it starts to improve or automate or uh whatever.

7:15 That's right.

7:15 And so so the weird thing is like there there is a question

7:18 of is it different this time in terms of the jobs because it's

7:21 like if the software is fundamentally like human labor like equivalent then even

7:26 the business it creates won't the workers in that business also just be

7:30 AI labor mostly that's doing that right and do we

7:32 all just become sort of like mechanics for for the robots

7:36 and for uh for AI right that's one part

7:39 I haven't been able to wrap my head around

7:41 I haven't either and I haven't heard the case for why we do

7:43 that is compelling enough to Like the the most frequent example I hear is, "Hey,

7:49 in the 1950s there were skyscrapers full

7:52 of humans who were humanly calculating spreadsheets."

7:57 And then this then the real spreadsheet

7:58 came along and all those jobs are eliminated.

8:02 And then that's where the story ends for them.

8:04 And I'm like, "Hey, time out.

8:05 I was just in New York City.

8:06 Those skyscrapers aren't empty.

8:08 What happened?" And what happened was a plethora of things.

8:12 And I would use just one tiny little example of what

8:15 happened that created trillions and trillions of dollars of market cap.

8:20 Once an once a a spreadsheet was available that was digital, uh it could calc.

8:30 And so all of a sudden you could have

8:33 sophisticated pricing engines like black shores at your fingertips.

8:38 And that meant that you could now price

8:40 options and futures and derivatives that didn't exist before.

8:44 And so entire markets got created on top of that one

8:47 technological change that employ hundreds of thousands of people.

8:52 Um, and again that was unknowable on the on the other side

8:56 when the spreadsheet was just being invented that somebody would deploy blacks

9:00 and somebody would figure out how to price puts and calls

9:03 and derivatives and make entire markets

9:06 like the options exchange, the merc, etc.

9:10 possible uh to be able to syndicate loans like

9:13 none of that stuff was available and now it's all

9:15 available and that's the kind of thing where I think

9:18 okay make me the case that all of a sudden

9:22 now the computers do all of that and take

9:25 all that creativity and all those jobs I just haven't

9:28 seen that in history and I don't know that I

9:30 can I can fully buy into that ending point.

9:34 What do you think happens with uh the different AI players?

9:38 You've got like you think Elon has called it

9:41 the highest ELO game in in the world, right?

9:43 So, it's the highest level chess game being played between the smartest people

9:46 with the most capital behind them going all out for the biggest prize.

9:50 You've got Zuckerberg and you've got Google and you've got Elon,

9:53 you've got Sam Alman, you've got Anthropic.

9:55 You have all these players.

9:57 How do you how do you see this playing out?

9:58 You've seen a couple of tech waves, you know, uh you know,

10:01 at least some of the players and maybe even all of them personally.

10:03 Um, how do you see this playing out?

10:07 I sort of wonder if if they're creating a app layer or a tooling layer.

10:13 And what I mean by that is I think they're creating the tooling layer

10:16 that then a lot of us are now using to create real businesses on top of.

10:22 So I kind of go back to the internet breakthrough

10:25 and the internet revolution and think about okay like the browsers everybody

10:30 was super excited about the browser business and Mark Andre made

10:34 uh made a career out of Netscape but I wonder if like

10:39 those now browsers are a commodity but what got him what

10:43 got built on top of a browsable web were things like

10:48 Facebook and Airbnb and uh Erade and you name the the businesses

10:55 like tons of GDP got created on top of that tooling.

11:00 And it sort of feels like to me we're getting

11:01 really excited like we did about Netscape and Internet Explorer.

11:05 We're getting really excited about these hyperscalers,

11:07 but what you really want to be looking at is what

11:09 businesses are going to get created on top of this tooling.

11:12 And so I'm psyched for the work they're doing.

11:15 I think it's amazing at its breakthrough level,

11:17 but I but I am also much more excited

11:21 about what's going to get built on top of it.

11:23 Yeah, me too.

11:24 I can't wait to see it.

11:25 I almost want to fast forward.

11:26 It's like I don't want to miss it, but at the same time,

11:28 I'm so curious how the how this all plays out,

11:31 how the what the world looks like in 10 years.

11:32 This is probably the first time that I feel like I I

11:37 have no idea what the world's going to look like in 1015 years,

11:40 but I do know it's it's likely to be

11:42 very like completely different, you know, for sure.

11:45 You know, like I I bought a Tesla

11:47 this year and it has like self-driving that's perfect.

11:50 It's amazing.

11:51 It's it's so good and I just don't drive anymore.

11:54 And I'm like trying to tell my mom and my sister and I'm just like,

11:57 "Guys, you realize like you don't have to drive anymore?

11:59 Like that's that's now." And I I can't believe that that's

12:02 not more people aren't freaking out about this one change,

12:05 let alone whether it's robots or it's AI

12:08 that you know AI assistance in everybody's pocket

12:10 or totally there's so many different ways that I

12:12 feel like the whole world in 1015 years

12:16 is going to like my kids won't really understand

12:20 wow you used to have to do all that you used to do that like

12:22 that's so funny that you used to do that's horseback riding right not cars

12:26 that and it seems like we're on one of those transition points which exciting

12:30 totally agree