Shadowing practice: AI Sells Labor, Not Software — Legendary Investor Elad Gil

Listen to the line, then say it out loud
Line 1

Full transcript

  1. I'm looking at a piece in front of me.0:00
  2. This is from a while ago, but it's you discussing long held dogma that ends up being unbiable.0:02
  3. So for instance, the common held belief after PayPal sale to eBay that fraud will kill you in the payment space.0:11
  4. Right?0:17
  5. Yeah.0:17
  6. And I'm wondering how you orient yourself as an investor to stress test those types of dogma.0:17
  7. It's really hard because you often end up, you start off with some set of beliefs.0:26
  8. You think something's interesting.0:31
  9. Well, maybe you invest in it, maybe you start a company in it.0:33
  10. And then it turns out that thing you think is really interesting turns out to be really hard and you get killed.0:36
  11. And then five years later, a company comes up that actually does it and wins.0:40
  12. And the question is why?0:45
Show all 122 lines Show fewer lines
  1. Why did the thing suddenly work when it didn't before?0:48
  2. Or there's 10 attempts to do X and then suddenly, is it the technology got good enough?0:51
  3. It could be a regulatory change, it could be a market shift, it could be whatever.0:56
  4. An example that may be Harvey and legal where selling the law firms traditionally has been awful.0:59
  5. And Harvey's not much broader than that, right?1:04
  6. They also had very strong enterprise adoption and lots of different people using them in different ways.1:06
  7. But the dogma was always like building stuff for law firms as crappy as a business and you should never do it.1:11
  8. But what AI did is it shifted things from selling tools to selling work product or selling units of labor.1:15
  9. That's really the shift in generative AI.1:21
  10. We're going from seats and we're going from software and SaaS and we're moving into a world where we're selling human labor equivalents, we're selling work hours or labor hours1:24
  11. or whatever you want to call it, it's a cognition.1:34
  12. And so Harvey is effectively helping really augment lawyers in different ways.1:36
  13. And part of that's a knowledge corpus, but a lot of it is this tooling that really helps lawyers achieve the goals that they have in different ways in a1:41
  14. collaborative manner in some cases.1:47
  15. And so this is a fundamentally different type of product from what people were selling before.1:48
  16. And so it opened up the market in a way that the market wasn't open before.1:52
  17. There's actually a broader conversation around is the world market limited or founder limited in terms of entrepreneurial success.1:55
  18. The way I culminate our school of thought is that we just don't have enough founders and if we had 10 tens of many founders, we'd have 10 tens of2:03
  19. many big companies.2:08
  20. And there's an alternate school of thought, which is how many markets are actually open in any given moment in time.2:10
  21. And those are the ones where you can build big companies because if the market isn't open to innovation or change or whatever or hasn't is undergoing a shift, you2:15
  22. can't really build anything or any house.2:22
  23. So why do it?2:24
  24. And the striking thing about AI is it's opened up tons and tons of markets that were closed for a long time and it's opened it up because of capabilities.2:25
  25. But it's also opened it up because every CEO is asking themselves what's my AI story and the worry more open us to try things that I've ever seen in2:33
  26. my life.2:41
  27. And so we have this odd moment in time where things are massively available for founders to do new things.2:41
  28. And if you're an AI company and you're not seeing explosive growth quickly, something's fundamentally broken because the markets are so open that you can suddenly grow at a rate2:48
  29. that you've never grown before.2:58
  30. There's always been cases of companies that just go like this.2:59
  31. But again, you look at the ramps of open and anthropocon, it's the fastest ramps to tens of billions ever percentages of GDP.3:02
  32. It's like crazy.3:08
  33. If we come back to your comment of not necessarily market first and strength of team second all the time, but like you said, you 90 % agree with that,3:08
  34. right?3:18
  35. And if you have an excellent team in a terrible market, that's going to be a difficult one to execute.3:19
  36. How do you determine what is a good versus great market or just what is a great market?3:26
  37. What do you look for?3:31
  38. And the example you gave, I might be over reading this, but when you said that when Google shut down, I think it was Maven, right?3:32
  39. That's an interesting kind of event based approach as an input to investing, right?3:40
  40. Because you're like, okay, if they're not going to build it, that suddenly creates a playing field for startups to play in that space.3:45
  41. So could you speak to more of how you determine or look for great markets?3:56
  42. I mean, there's a few different ways to think about it.4:00
  43. One is like, some people take the framework of why now, what's shifted now that makes it something an interesting market because people have been trying to do things for4:02
  44. a long time in every market.4:09
  45. And so that may be a regulatory shift, right?4:10
  46. Some SARA, the fleet management company benefited from the fact that some of those regulation around needing in -cap monitoring of drivers.4:13
  47. So you had suddenly cameras watching people so they don't fall asleep while they're driving trucks on the road, right?4:19
  48. And so that was another entry point at that start building, not a suite of software.4:24
  49. But it was a regulatory shift.4:27
  50. Sometimes there's technology shifts like what's happening in AI.4:28
  51. And the crazy thing about the AI shift is that foundation models instantly plugged into a massive set of markets, which is basically all enterprise data and information and email4:31
  52. and just all -way color work was suddenly available to AI because it was the perfect technology for that.4:43
  53. It also plugged into code, which is a type of white color work.4:49
  54. So suddenly it just inserts into language and languages used everywhere in enterprises as well as in consumers.4:51
  55. So there's just a massive market to tap into and transform a set of markets.4:56
  56. Robotics is a little bit different from that because even if you had the world's best robotic model, the sub markets that already have robotic hardware are quite small on5:00
  57. a relative basis.5:06
  58. And so you don't have that instant runway that you would with language unless you come up with something new there.5:08
  59. That's kind of an aside.5:15
  60. But I think robotics is really interesting and will be important.5:16
  61. It's more just that nuance of like, what's that instant thing you plug into commercially?5:18
  62. And then there's regulatory shifts, there's technology shifts, there's incumbency or company shifts, competitive shifts.5:22
  63. A company may blow itself up and may get bought by a competitor.5:29
  64. One company I'm excited about on the security side is called InPhysical and they're basically competing in part with Hashi.5:32
  65. Hashi got bought by IBM.5:38
  66. Anytime you get bought by IBM, you slow down a lot usually.5:39
  67. Certainly it creates more opportunity for a startup.5:42
  68. So I just feel like there are these different things that can change at a given moment in time.5:44
  69. It could be the market's growing really fast.5:49
  70. That's Coinbase and crypto, right?5:50
  71. You just have something of this adoption and proliferation of token types.5:52
  72. There's lots and lots and lots of different markets that are interesting.5:55
  73. The commonality is usually like, is it also big?5:58
  74. Is there a big enough TAM and there's two types of TAMs?6:01
  75. There's fake TAM.6:03
  76. Just for people listening who might not have it, a total addressable market.6:03
  77. Probably a best market.6:07
  78. So what's the market you're in?6:07
  79. And sometimes people come up with these fake markets.6:09
  80. They're like, oh, well, we are facilitating global e -commerce and global e -commerce.6:11
  81. I'm making up the numbers.6:18
  82. $30 trillion a year.6:19
  83. And so we're in a $30 trillion a year market.6:20
  84. And if we get just a tenth of a percent of that, it's $300 billion of revenue.6:22
  85. That's not your market.6:25
  86. Your market is like, you built this little optimization engine for SMB websites or whatever.6:28
  87. That's not a $30 trillion market.6:33
  88. And so really, it's kind of defining the market.6:36
  89. There's a really famous example of this.6:38
  90. We're defining your market changes.6:40
  91. How you think about it.6:42
  92. And so that's Coca -Cola, right?6:43
  93. So Coke and Pepsi were roughly neck and neck in terms of market share for decades.6:45
  94. And then one of the Coke CEOs said, hey, maybe we should be thinking about our share is share of liquid sold, like drinks, not share of soda.6:50
  95. And so we just went from 50 % market share to 0 .5%.7:02
  96. And that's why they bought the Sony and that's why they entered all these other markets, because they said our definition of our market is wrong.7:06
  97. We're out of the set of pop business, we're in the drinks business.7:13
  98. And so I think also sometimes reconceptualizing what you're doing can really help change your scope of ambition or how you think about what you're doing.7:15
  99. If you were trying to spot along the lines of the fraud will kill you in the payment space, any dogma in the AI world, the sphere of AI, anything7:22
  100. hop to mind where you think, maybe that's not true now, or maybe in like two years, it'll be completely untrue, but people will have latched onto this belief as7:35
  101. one of the thou shalt not or thou shalt commandments.7:46
  102. I don't know.7:52
  103. I mean, there's some things that have circulated in the past around what's the ROI and the capex spend of the whatever be paid back.7:52
  104. And I just like, I think that stuff is probably off.7:58
  105. But yeah, I think fundamentally, there are moments in time where it's very smart to be contrarian.8:01
  106. And there are moments in time where being consensus is the smartest possible thing you can do.8:06
  107. And I think right now we're in a moment in time where being consensus is very right.8:10
  108. You can really overthink it and what's a contrarian thing, we should go do a bunch of hardware stuff because blah, blah, blah.8:15
  109. You know, it may just buy more AI, you know what I mean?8:21
  110. I think people make these things way too complicated.8:23

American accent training from a fast, unscripted tech conversation

This clip is english speaking practice built from a real interview, not a script written to be read aloud. It is a segment of The Tim Ferriss Show, where Ferriss asks investor Elad Gil how he tests long held business beliefs, and Gil answers with the example of Harvey, a company selling AI tools to law firms. The recording sits at C1 and runs at 224 words per minute, faster than 81 percent of the catalogue, so the sentences arrive close together with almost no pause between them.

It is american accent training precisely because nothing here is simplified for learners. Gil talks the way people actually think out loud: he starts a sentence, corrects the direction mid way, and lands on the point a few clauses later. Words like optimization, entrepreneurial, regulatory and collaborative show up inside ordinary sentences rather than as vocabulary flashcards, and the conversation format means you get two different voices, tempos and turn taking patterns in one clip, not one narrator reading at a fixed rhythm.

The language in this exchange: dogma, labor and Harvey

Three lines carry most of the interesting language in the opening of this clip. Quote them out loud a few times before you try to say them at speed.

  • "So for instance, the common held belief after PayPal sale to eBay that fraud will kill you in the payment space." This is one long noun phrase stacked on another with no comma to breathe on. Say it in chunks the first few times: "the common held belief", then "after PayPal sale to eBay", then the rest, before joining them.
  • "But the dogma was always like building stuff for law firms as crappy as a business and you should never do it." The word "like" here is filler, not a comparison, and "crappy" carries the stress. Practising this line is a good check on whether you can hear casual filler words without losing the sentence.
  • "But what AI did is it shifted things from selling tools to selling work product or selling units of labor." Three parallel verb phrases in a row, each starting with "selling". Say them as a rising list, with the stress falling on "tools", "product" and "labor" in turn, the way Gil does.

How to shadow this C1 clip line by line

This is english listening practice as much as speaking practice, because at 224 words per minute you have to hear the sentence boundary before you can repeat it. Work through the opening in this order.

  • Play lines 1 through 6 once with no pausing, just to catch the question and answer rhythm between the two speakers.
  • Isolate line 3, the PayPal and eBay sentence, and shadow it five times before moving on. It is the longest single clause in the opening and the best test of whether you can keep pace.
  • Move to lines 7 through 13, where Gil lays out his general point about dogma. These are shorter sentences, good for building speed after the long one.
  • Slow down for lines 16 to 20, the Harvey example, and shadow this line by line rather than as one block: the argument only makes sense if you keep the order of the clauses.
  • Finish on line 22, the longest sentence in the excerpt, and repeat it until the three "selling" phrases come out evenly spaced. This is where daily english conversation practice pays off, since real conversation rarely gives you a tidy short sentence to imitate.
  • Once you can keep up with lines 1 to 22 without the audio running away from you, you have a working method for how to pronounce english words in this kind of fast, argument by example American speech, and a real basis to judge whether your own pace is improving.
Download on the App Store