WEBVTT

1
00:00:00.000 --> 00:00:07.396
<v Alex>Last Thursday my board asked which AI model we'd picked. One word, they said. I said "all of them," and nobody loved it.

2
00:00:07.396 --> 00:00:11.898
<v Sam>Which is the right answer, and also the answer that makes a board nervous.

3
00:00:11.898 --> 00:00:18.973
<v Alex>Right. And on the way home I realised the model isn't the interesting bit. Whoever decides which model gets which job is.

4
00:00:18.973 --> 00:00:24.118
<v Sam>Hm. I'd have said the opposite in that room. The model is what everybody argues about.

5
00:00:24.118 --> 00:00:36.981
<v Alex>Today's story says that argument is about to look quaint. The company is Perplexity, which has never trained a frontier model and whose revenue run rate has quadrupled this year. Perplexity: the AI prize is the switch, not the model.

6
00:00:36.981 --> 00:00:56.919
<v Alex>Before we get into it, one thing from my week. My head of payments forwarded me a headline on Monday: OpenAI's new wallet taps into the card networks. That's all I've got, a headline, no details yet. And my first reaction was, great, now an AI can buy things from my stores, and my store has no idea who it's talking to.

7
00:00:56.919 --> 00:00:59.170
<v Sam>Because to you it's a security problem.

8
00:00:59.170 --> 00:01:04.637
<v Alex>Everything's a security problem to me. My head of security says I've started to sound like her.

9
00:01:04.637 --> 00:01:06.566
<v Sam>I'd pay to hear that conversation.

10
00:01:06.566 --> 00:01:08.817
<v Alex>She did not laugh. What about you?

11
00:01:08.817 --> 00:01:22.002
<v Sam>Mine is the same headline from the other side. Our customers mostly buy the same bag of dog food every month. If a wallet inside a chatbot starts doing that reordering for them, who owns the relationship? Us, or the chatbot?

12
00:01:22.002 --> 00:01:25.539
<v Alex>And that's today's episode in miniature. Who sits in the middle?

13
00:01:25.539 --> 00:01:27.790
<v Sam>Then let's go sit in the middle.

14
00:01:27.790 --> 00:01:44.834
<v Alex>Welcome to Dan's AI Intel, the show where we try to make sense of the fastest, most consequential shift any of us is likely to live through. I'm Alex, with Sam, and today is Perplexity again. If you like the show, follow it wherever you listen, so you don't miss the next one.

15
00:01:44.834 --> 00:01:50.622
<v Sam>And we've been here before. Perplexity was our episode two, back in April, and a lot has moved.

16
00:01:50.622 --> 00:02:06.380
<v Alex>Here's the map. Where we left it. Why the model is turning into the cheap part. How Perplexity moved its decision-making onto your own computer in a few months. How that works, whether the economics hold up, who else is building the same thing, and what could break it.

17
00:02:06.380 --> 00:02:11.846
<v Sam>And the question under all of it is a simple one. Who ends up holding the switch?

18
00:02:11.846 --> 00:02:35.000
<v Alex>Let me stake my claim, because I think I know. I think the owner of the device wins. Whoever owns the operating system sits closest to your files, your apps, your chip. I don't love saying it, because open usually beats closed when it works. But I said a long time ago that the company owning the whole stack has an advantage nobody beats, and I haven't seen a reason to move.

19
00:02:35.000 --> 00:03:00.867
<v Sam>And I'll end up on the other side. The neutral layer wins, because it's the only one sitting on my side of the table. Every time I've been locked into one vendor, the price went up the following year. My renewals are the proof. When I asked one vendor for a cap on usage, they told me it was usage-based, which is a polite way of saying no. So hand me the thing that can switch suppliers when mine misbehaves.

20
00:03:00.867 --> 00:03:03.454
<v Alex>You're saying it because lock-in is your nightmare.

21
00:03:03.454 --> 00:03:06.687
<v Sam>I'm saying it because I sign the budget. Different thing.

22
00:03:06.687 --> 00:03:08.951
<v Alex>We'll see where we both end up.

23
00:03:08.951 --> 00:03:16.388
<v Sam>So, quick recap for people who missed episode two, which came out in April. Alex, you do it, because I'll mix something up.

24
00:03:16.388 --> 00:03:38.375
<v Alex>You will. So, three things. Perplexity started as an answer engine. It cited its sources and it set itself up against Google search. Second, the pressure of fighting Google on search alone pushed it toward something bigger. It went from finding information to doing the work. That's Computer, which launched in February: you hand it a goal and it works through it, including operating the web for you.

25
00:03:38.375 --> 00:03:52.602
<v Sam>What I took from April was the shift from answers to actions. I stopped being impressed by answers around the time I'd rather it just did the thing, and then I stopped being impressed by that the first time it did the wrong thing.

26
00:03:52.602 --> 00:03:54.542
<v Alex>That's the right scar to have.

27
00:03:54.542 --> 00:04:02.625
<v Sam>And the open question at the end was whether that made it a better search tool or the early shape of a new operating system.

28
00:04:02.625 --> 00:04:17.176
<v Alex>Yes. And the operating system idea now has a specific shape. Computer was always a conductor, handing parts of a job to different models. Now it also decides where the work physically runs. In a data centre, or on the laptop in front of you.

29
00:04:17.176 --> 00:04:20.732
<v Sam>So episode two asked whether Perplexity was becoming an operating system.

30
00:04:20.732 --> 00:04:27.846
<v Alex>And this one says which part of an operating system it's actually building. The scheduler. The piece that decides what runs where.

31
00:04:27.846 --> 00:04:29.786
<v Sam>That's the part nobody ever sees.

32
00:04:29.786 --> 00:04:36.899
<v Alex>Which is what I love about it. Nobody has ever sent a thank-you note to a scheduler, and nothing works without one.

33
00:04:36.899 --> 00:04:39.809
<v Sam>You're about to go on a tangent, aren't you.

34
00:04:39.809 --> 00:04:40.779
<v Alex>A short one.

35
00:04:40.779 --> 00:04:41.103
<v Sam>Alex.

36
00:04:41.103 --> 00:04:42.396
<v Alex>Fine. I'll save it.

37
00:04:42.396 --> 00:04:50.480
<v Sam>Start me at the claim I resist. The model is the cheap part. In my world the model is the line item that keeps growing.

38
00:04:50.480 --> 00:05:06.000
<v Alex>It grows on your invoice. It's falling underneath. Our bubble episode, number 49, from late August, said the exposed layer of the market is the lab whose only product is a model. They rent their compute, and they price against open models that cost almost nothing to run.

39
00:05:06.000 --> 00:05:08.129
<v Sam>And Kimi K3, number 33, in July.

40
00:05:08.129 --> 00:05:25.163
<v Alex>The Chinese lab that got to the frontier and then handed its weights to anyone who wanted to run them. And the Apple one, number 30, also July, where we said compute is splitting in two. A thin, expensive frontier for the hardest thinking, and a fast-falling cheap tier that runs on hardware people already own.

41
00:05:25.163 --> 00:05:31.246
<v Sam>So the story we told in the summer was the market. Today's is a company built to live inside it.

42
00:05:31.246 --> 00:05:40.979
<v Alex>That's it exactly. Epoch AI, who track capability, found in May that since the start of the year the best open models have trailed the best closed ones by about four months.

43
00:05:40.979 --> 00:05:55.579
<v Sam>Four months. That's a season. And my contracts don't know what a season is. I signed my team's tools for twelve months in January, and some of that is already overpriced for what it does. If the lead is four months, why am I locked in for twelve?

44
00:05:55.579 --> 00:06:02.575
<v Alex>And Epoch's tracking of prices says the cost of reaching a fixed level of capability falls by something like forty times a year.

45
00:06:02.575 --> 00:06:09.571
<v Sam>Forty. Per year. Okay, but that's not what I see. What I see is my vendor telling me the new model costs more.

46
00:06:09.571 --> 00:06:13.525
<v Alex>The newest one, sure. The same capability a year later costs a sliver.

47
00:06:13.525 --> 00:06:15.958
<v Sam>So I'm paying a premium to be early.

48
00:06:15.958 --> 00:06:27.821
<v Alex>You're paying a premium to be first, and your team might not need first. Most steps inside a long task don't need the biggest model. That's the whole insight. A task is thirty small steps and two hard ones.

49
00:06:27.821 --> 00:06:31.167
<v Sam>And right now I pay the hard-step price for all thirty-two.

50
00:06:31.167 --> 00:06:40.596
<v Alex>At my company too. Our engineers point the best model at everything because it's the default and nobody wants to be the person who picked the cheap one and got blamed.

51
00:06:40.596 --> 00:06:43.333
<v Sam>Nobody ever got fired for buying the expensive model.

52
00:06:43.333 --> 00:06:47.288
<v Alex>That's the sentence my CFO is going to quote at me in January.

53
00:06:47.288 --> 00:07:01.583
<v Alex>Now, none of this says frontier models are worthless. The hardest reasoning still needs the biggest models, and the labs that build them are still the highest-valued private companies in the industry. The claim is smaller. A single model is rarely the whole answer to a task.

54
00:07:01.583 --> 00:07:04.625
<v Sam>And Perplexity's boss, Aravind Srinivas, has built everything on that.

55
00:07:04.625 --> 00:07:14.358
<v Alex>He told Fortune, the day after Computer launched, "The orchestration is the product. The model is a tool." And he added that the model layer is the most competitive it's ever been.

56
00:07:14.358 --> 00:07:19.225
<v Sam>A man selling a product that depends on other people's models saying models are tools. Convenient.

57
00:07:19.225 --> 00:07:32.000
<v Alex>Convenient, and consistent with what he built. At launch Computer orchestrated nineteen models. Claude Opus 4.6 from Anthropic did the orchestrating and the coding, Google's Gemini the deep research, Grok from xAI the quick jobs, and GPT-5.2 from OpenAI the long-context recall.

58
00:07:32.000 --> 00:07:36.091
<v Sam>Four labs that compete with each other, doing jobs for one company.

59
00:07:36.091 --> 00:07:43.591
<v Alex>He compared it to hiring. "When you build a team, you don't build a homogenous group where everyone has the same skills."

60
00:07:43.591 --> 00:08:05.750
<v Sam>That one I buy, though there's a catch. When I hire a specialist, I also get their judgement about when to stay out of a project. A router picks a model for a step. It doesn't know whether the step was worth taking. Somebody still decides that, and she's on my payroll. Also, my creative group would be a disaster if everyone wrote like me.

61
00:08:05.750 --> 00:08:08.477
<v Alex>Your copywriters would be thrilled to hear that.

62
00:08:08.477 --> 00:08:11.545
<v Sam>They'd be less thrilled by how much I edit.

63
00:08:11.545 --> 00:08:24.500
<v Alex>So here's where I want to plant a flag, and I know it's where we'll disagree. If the models are suppliers competing on price and a four-month lead, the durable seat is the buyer who chooses among them.

64
00:08:24.500 --> 00:08:27.227
<v Sam>Which is exactly my side. The neutral buyer.

65
00:08:27.227 --> 00:08:30.977
<v Alex>Or the buyer who owns the building the suppliers come to.

66
00:08:30.977 --> 00:08:33.705
<v Sam>That's a very Alex way of saying "me."

67
00:08:33.705 --> 00:08:34.727
<v Alex>I deserved that.

68
00:08:34.727 --> 00:08:41.545
<v Sam>So tell me how it moved from a stage to something on a desk. I like things I can buy.

69
00:08:41.545 --> 00:08:53.477
<v Alex>This is the part that got me excited. It starts on a stage in June. During the Computex show in Taipei, in Intel's keynote, Srinivas demonstrated what Perplexity called the first hybrid local-server inference orchestrator.

70
00:08:53.477 --> 00:08:54.500
<v Sam>In English, please.

71
00:08:54.500 --> 00:09:05.750
<v Alex>A program working through confidential deal documents, with small models running on Intel laptop chips deciding what stays on the machine and what's safe to send to the big models in the cloud.

72
00:09:05.750 --> 00:09:07.795
<v Sam>So the laptop is the bouncer.

73
00:09:07.795 --> 00:09:17.682
<v Alex>The laptop is the bouncer. And the next day he told CNBC, and this is the line that frames it all, "The data center is coming to your laptop."

74
00:09:17.682 --> 00:09:22.455
<v Sam>Which is exactly what a company that doesn't own a data centre would say.

75
00:09:22.455 --> 00:09:25.182
<v Alex>Careful. We haven't got to that part yet.

76
00:09:25.182 --> 00:09:28.591
<v Sam>So at Computer, sorry, at Computex, could I buy that?

77
00:09:28.591 --> 00:09:35.750
<v Alex>No. It was a demo, on Intel laptops, not for sale. And you did the Computer-Computex thing, which is your thing.

78
00:09:35.750 --> 00:09:38.818
<v Sam>It's a very similar word. Somebody should be fired.

79
00:09:38.818 --> 00:09:42.227
<v Alex>The product people, or the people who named the show?

80
00:09:42.227 --> 00:09:44.273
<v Sam>Both. Go on. What came after?

81
00:09:44.273 --> 00:10:00.295
<v Alex>Three products, in order. On the twenty-fifth of August, Portable Computer. The whole agent, its orchestrator and its models, sitting on a Nvidia DGX Spark desktop. An open model with twenty-seven billion parameters does the work, and the cloud only gets called when the user says yes.

82
00:10:00.295 --> 00:10:02.682
<v Sam>So that's the first one that's real.

83
00:10:02.682 --> 00:10:17.000
<v Alex>And then on the first of September, hybrid compute came to the Mac. A task can start in the cloud, where a frontier model plans and searches and reasons, and then hand the private-file steps to a model running on the Mac.

84
00:10:17.000 --> 00:10:19.098
<v Sam>That one I could actually have.

85
00:10:19.098 --> 00:10:30.285
<v Alex>If you've got an Apple silicon Mac with enough memory, yes. And on the twentieth of October, the fully local mode. The model, the agent and the execution all on the machine.

86
00:10:30.285 --> 00:10:40.423
<v Sam>Three products in under five months. What I notice is that every one is sold on keeping things private. Fear sells first. Savings come later, once people trust it.

87
00:10:40.423 --> 00:10:49.862
<v Alex>Fear sells first is right, and it's my CISO's whole worry. Privacy was the easiest thing to prove, so it shipped first. Cost is the harder promise.

88
00:10:49.862 --> 00:10:56.154
<v Sam>Hm. That's a big claim from a man who builds things for a living and loves a roadmap.

89
00:10:56.154 --> 00:11:13.634
<v Alex>Around the router, the agent itself got more persistent. In September they added Automations, so Computer picks up recurring tasks on a schedule. And on the fourth of November, Teammates. Agents that own ongoing work, research, bug triage, a marketing campaign. It starts with people on the top plan, Max.

90
00:11:13.634 --> 00:11:28.667
<v Sam>A marketing campaign. Owned by an agent. My Director of Lifecycle runs our autoship messaging, and the first thing she'll ask isn't whether it's clever. She'll ask who's on the hook when it sends a thousand wrong emails to people who already cancelled.

91
00:11:28.667 --> 00:11:44.398
<v Alex>Owned by one. Now, why does that matter for routing? Because a persistent agent multiplies the decisions. One that works every day for weeks makes thousands of small choices about what runs where, and most of those are cheap, repetitive ones that suit local hardware.

92
00:11:44.398 --> 00:11:54.187
<v Sam>So the router matters more the longer the agent lives. And that's what I'd budget for. Not the single clever task. The thousand dull ones I forgot about.

93
00:11:54.187 --> 00:12:03.276
<v Alex>Then I'd route my store-systems chatter locally first, the inventory checks and order lookups nobody watches. That's where a bill hides, and the data stays inside.

94
00:12:03.276 --> 00:12:05.724
<v Sam>And two more pieces landed this week.

95
00:12:05.724 --> 00:12:15.862
<v Alex>This week, yes. Twelve November, Decider, which we'll come to. And on the sixteenth Srinivas told Semafor that Perplexity had crossed a billion dollars in annual revenue run rate.

96
00:12:15.862 --> 00:12:19.358
<v Sam>A billion. That's a company that quadrupled in a year.

97
00:12:19.358 --> 00:12:34.390
<v Alex>It was under two hundred and fifty million when the year began, going by The Information's reporting. And the same day they announced a research lab, Perplexity Frontier Research, with more than ten thousand GPUs. Hold on to that one. It comes back.

98
00:12:34.390 --> 00:12:36.488
<v Sam>You're teasing me with a loop.

99
00:12:36.488 --> 00:12:37.886
<v Alex>I'm keeping you listening.

100
00:12:37.886 --> 00:12:40.683
<v Sam>It's working, annoyingly. I'm already waiting for it.

101
00:12:40.683 --> 00:12:45.577
<v Sam>Right, so now you get to be nerdy. How does a task actually move?

102
00:12:45.577 --> 00:13:09.000
<v Alex>Let's take hybrid compute on the Mac, because it's the version most people could use. Picture two rooms. In the cloud room, a frontier model does what needs a frontier model: it plans the task, searches the web, does the heavy reasoning. In the Mac room, a smaller model does what shouldn't leave the machine. Reading your private files, handling sensitive details, acting inside your own apps.

103
00:13:09.000 --> 00:13:11.513
<v Sam>And in the doorway between the rooms?

104
00:13:11.513 --> 00:13:30.183
<v Alex>A gate. Perplexity calls it a privacy gate. A classifier on the device looks at each step before anything leaves, and it flags names, addresses, account numbers and secrets. Then it does one of four things. It masks the detail, keeps the step local, refuses the action, or asks you for consent.

105
00:13:30.183 --> 00:13:34.132
<v Sam>So the little model is basically a bouncer holding a clipboard.

106
00:13:34.132 --> 00:13:46.698
<v Alex>With a clipboard and a rule book. Credentials, card numbers and government ID numbers get the strictest treatment. And enterprise administrators can set rules for a whole organisation and audit when information leaves a device.

107
00:13:46.698 --> 00:14:09.676
<v Sam>Audit. Now that I like. That's a word my finance team understands. But four options is a menu, not a test. My customer service inbox is the test: a name, an address and half a card number in one angry message at eleven at night. Does it mask the card and keep the anger? Because the anger is the part I need to read.

108
00:14:09.676 --> 00:14:26.551
<v Alex>And my CISO. She stopped me in the corridor this week and asked what that gate would flag before any agent touches our store customer files. I said I didn't know. She said, find out, and don't tell me it's open source as though that's an answer.

109
00:14:26.551 --> 00:14:27.987
<v Sam>Is it open source?

110
00:14:27.987 --> 00:14:37.681
<v Alex>The classifier is. A model of about six hundred million parameters, small enough to run on any machine that can run the agent. Perplexity gave it away.

111
00:14:37.681 --> 00:14:40.553
<v Sam>Why would a company give away the gate?

112
00:14:40.553 --> 00:14:53.837
<v Alex>That's the bit I'd underline. The decision point for what leaves the device is exactly what they open-sourced. You only do that if you think the money is in the system around it, not the part itself.

113
00:14:53.837 --> 00:14:59.581
<v Sam>Or you do it so outsiders can test it and your lawyers can say they did.

114
00:14:59.581 --> 00:15:01.018
<v Alex>Both can be true.

115
00:15:01.018 --> 00:15:02.095
<v Sam>Mine usually are.

116
00:15:02.095 --> 00:15:09.993
<v Sam>And the model on the Mac, whose is it? Please say it's theirs, so I know whose neck my files are in.

117
00:15:09.993 --> 00:15:21.841
<v Alex>Not theirs. The local models are open-weight models from other labs, sometimes retrained by Perplexity. Hybrid compute launched with Google's Gemma 4 E4B, Alibaba's Qwen3.6 35B-A3B, and a Perplexity model post-trained for Computer.

118
00:15:21.841 --> 00:15:23.637
<v Sam>So, Google on my laptop.

119
00:15:23.637 --> 00:15:27.945
<v Alex>Gemma. Not Gemini, which stays in the cloud. Same company, different animal.

120
00:15:27.945 --> 00:15:29.740
<v Sam>Hardware, though. What's the bar?

121
00:15:29.740 --> 00:15:39.434
<v Alex>Real. Hybrid compute wants an Apple silicon Mac with at least twenty-four gigabytes of memory, and the fully local mode wants more. That's a well-equipped professional's laptop.

122
00:15:39.434 --> 00:15:52.000
<v Sam>So the people who can use this are the people with the expensive laptops. Which leaves out my copywriters on their older machines, and they're the ones who'd gain most from work that costs nothing.

123
00:15:52.000 --> 00:15:57.187
<v Alex>Which are also, as it happens, the people most likely to be handling confidential files.

124
00:15:57.187 --> 00:16:02.029
<v Sam>Hm. My son has a gaming laptop with a model on it right now.

125
00:16:02.029 --> 00:16:04.449
<v Alex>Theo has a model on his laptop?

126
00:16:04.449 --> 00:16:17.936
<v Sam>A small one, for his Minecraft mods. No account, no cloud, and he's very smug about it. He told me the data never leaves the machine, and I realised that's roughly the pitch I'm being sold for a premium.

127
00:16:17.936 --> 00:16:20.357
<v Alex>A fourteen-year-old who invented the privacy gate.

128
00:16:20.357 --> 00:16:24.852
<v Sam>He didn't invent the gate. He just doesn't have anything private to protect.

129
00:16:24.852 --> 00:16:29.002
<v Alex>That's the most accurate review of a privacy feature I've ever heard.

130
00:16:29.002 --> 00:16:39.377
<v Alex>Decider is the newest piece, and I think it's the revealing one. It's a family of small open-weight models, published twelve November, and they don't write text. They return probabilities.

131
00:16:39.377 --> 00:16:44.564
<v Sam>Meaning what? Give me the version I could explain at a stand-up without a diagram.

132
00:16:44.564 --> 00:17:01.163
<v Alex>You ask it a question about some content, and instead of an essay it hands back a score for each possible answer. Is this a search question, yes or no? How deep should the research go? Should this step go to a smaller model or a bigger one?

133
00:17:01.163 --> 00:17:18.108
<v Sam>So it's a sorting hat, and a quick one. Which means everything depends on the sort being right, and I've seen what a wrong sort costs. Our own targeting once filed a lapsed customer as loyal, and we sent her a thank-you for her loyalty the week she cancelled.

134
00:17:18.108 --> 00:17:35.053
<v Alex>A router's brain. Perplexity says it already runs a classifier like this on every query. One head decides whether a search is needed, another how deep to go. The larger Decider, built on a Qwen model, beat the classifiers it runs in production by a bit over six points.

135
00:17:35.053 --> 00:17:40.240
<v Sam>Slow down. I can feel my eyes doing the thing they do in budget reviews.

136
00:17:40.240 --> 00:17:45.081
<v Alex>Fair. It beats what they use now, and both sizes are free to download.

137
00:17:45.081 --> 00:17:49.923
<v Sam>And this connects to Jev, doesn't it. The fast one from that September episode.

138
00:17:49.923 --> 00:17:54.073
<v Alex>Number sixty-one, the fast text-free one. Perplexity benchmarks Decider against Jev directly.

139
00:17:54.073 --> 00:18:06.868
<v Sam>So the router's brain is turning into its own category. That surprises me. My martech vendors will start selling that brain as a separate line item, and I can already see the invoice I'll have to defend.

140
00:18:06.868 --> 00:18:15.859
<v Alex>It's turning into a product. Across everything we've covered, Perplexity builds the small fast parts that decide, and buys or borrows the large parts that think.

141
00:18:15.859 --> 00:18:29.000
<v Sam>A company whose edge is the harness and the classifiers, not a frontier model. That's a strange thing to build a billion-dollar run rate on, and it's the kind of dependence I flag in my own agency contracts.

142
00:18:29.000 --> 00:18:39.395
<v Alex>The harness. My word of the month. And here's what it means for me on Monday. I should own the piece of my stack that decides, and rent the piece that thinks.

143
00:18:39.395 --> 00:18:43.618
<v Sam>Which sounds good until you ask who's responsible when the decider is wrong.

144
00:18:43.618 --> 00:18:59.210
<v Alex>So here's the thesis, and then I want you to poke it. The idea that the winning model is a smart router across cloud and edge is right on direction. Today it routes mainly on privacy and on what the user permits. It doesn't yet route on cost.

145
00:18:59.210 --> 00:19:08.631
<v Sam>Say that again, slowly, because that's the sentence I'd be paying for. If it routes on privacy and not on price, then I'm still the one doing the weighing.

146
00:19:08.631 --> 00:19:24.223
<v Alex>Start with what Srinivas says the router is for. He told CNBC one metric will decide the winner: whoever delivers the most token value per watt per user. Winning means, in his words, balancing accuracy, latency, cost, privacy and intelligence all together, and "this is an orchestration problem."

147
00:19:24.223 --> 00:19:34.618
<v Sam>Five things at once. That's a lot to ask of anything. I've watched my own team fail at two. Accuracy and cost alone start a fight in every planning meeting I run.

148
00:19:34.618 --> 00:19:39.815
<v Alex>And he called the product a hybrid neutral orchestration layer, and the most versatile operating system.

149
00:19:39.815 --> 00:19:45.662
<v Sam>And that's the scheduler line you kept stopping yourself saying, so go on, you've clearly been dying to.

150
00:19:45.662 --> 00:20:12.624
<v Alex>Yes! Okay, I'm allowed now. In a normal operating system, the scheduler decides which program runs on which processor core and when. You never see it and nothing works without it. Perplexity's router does the same job one level up. The programs are the steps of a task, the processors are models and machines, and the rules weigh those five things. An app that owns that decision is as close to an operating system as an app can get without owning the device.

151
00:20:12.624 --> 00:20:17.172
<v Sam>Is that the tangent? I braced for twenty minutes and that was under one.

152
00:20:17.172 --> 00:20:18.471
<v Alex>That was the tangent.

153
00:20:18.471 --> 00:20:23.669
<v Sam>It was good, and I hate that it was good, because now you'll do it again.

154
00:20:23.669 --> 00:20:24.318
<v Alex>Thank you.

155
00:20:24.318 --> 00:20:30.490
<v Sam>And now I'd like to be unfair to it, because that's my job on this show. What ships today?

156
00:20:30.490 --> 00:20:43.809
<v Alex>In hybrid compute, you open the model picker, choose hybrid, and pick both the local and the cloud model yourself. The classifier decides which steps are private. The other two run locally and reach the cloud only when you authorise it.

157
00:20:43.809 --> 00:20:49.006
<v Sam>So you pick the models. Then who's doing the routing, exactly, if I'm the one choosing?

158
00:20:49.006 --> 00:21:02.000
<v Alex>You pick the models. All three products route on one thing, how sensitive the data is, and you set the rest. The automatic weighing of cost and speed against accuracy is the direction, not the behaviour you can buy today.

159
00:21:02.000 --> 00:21:08.667
<v Sam>So it's a smart router that asks you what to do. I hired that person once. It didn't end well.

160
00:21:08.667 --> 00:21:12.000
<v Alex>Today it's a very good bouncer and a hopeful scheduler.

161
00:21:12.000 --> 00:21:15.667
<v Sam>So I wait for the version that actually does the weighing.

162
00:21:15.667 --> 00:21:20.667
<v Alex>Ship it. That's what I tell my own team, and Perplexity should hear it too.

163
00:21:20.667 --> 00:21:22.667
<v Sam>Okay, but who pays for that?

164
00:21:22.667 --> 00:21:41.667
<v Alex>Quietly, Perplexity does less than you'd think. Here's the mechanism people miss. They say plainly that on-device work doesn't consume credits, and that local-first means people no longer have to ration what they use intelligence for. Every step that runs on your Mac or your DGX Spark is a step Perplexity doesn't pay a cloud provider for.

165
00:21:41.667 --> 00:21:53.333
<v Sam>So when it moves my work onto my machine, my machine pays for it. And my electricity. Which is fine for one customer, and a different conversation when you're buying machines for a whole team.

166
00:21:53.333 --> 00:22:16.333
<v Alex>Your hardware, your electricity, their margin. A router that moves work onto machines customers already own lowers the cost of serving a subscriber. It matters because they buy most of their intelligence from other labs. And Srinivas more or less said so in June, when he attributed the tripling of their revenue run rate in the first five months of the year to the model advances made by Anthropic.

167
00:22:16.333 --> 00:22:27.000
<v Sam>So their growth leans on the labs they route to. Which is a familiar position. My agencies grow when my budget grows, and they're lovely to me right up until the renewal.

168
00:22:27.000 --> 00:22:28.667
<v Alex>Which is exactly the exposure.

169
00:22:28.667 --> 00:22:46.333
<v Sam>This is the bit that's landed on me this week, actually. My friend Lauren, who runs a six-chair hair salon, got a note that her booking software is moving to credits. She called me and asked what a credit is, and I said, honestly, I don't know, and neither does the vendor yet.

170
00:22:46.333 --> 00:22:48.000
<v Alex>And you assumed the worst.

171
00:22:48.000 --> 00:22:49.667
<v Sam>It's my area of expertise.

172
00:22:49.667 --> 00:22:51.667
<v Alex>Credits are tokens in a wig.

173
00:22:51.667 --> 00:22:53.333
<v Sam>That's going straight to her.

174
00:22:53.333 --> 00:23:03.333
<v Alex>But look at what Perplexity's doing about the other side of it. "Doesn't consume credits" is a promise to Lauren-shaped customers. It says: the local step is free to you.

175
00:23:03.333 --> 00:23:08.000
<v Sam>And I'd believe that more if I'd seen the router choose it for me.

176
00:23:08.000 --> 00:23:11.333
<v Alex>Which is the same thing you said about the price.

177
00:23:11.333 --> 00:23:28.333
<v Alex>One more thing, and it's the loop I opened. There's a signal pulling the other way. Perplexity Frontier Research, announced alongside the billion-dollar milestone, lists pre-training among its pillars and promises to release models. A company that really thought models were interchangeable parts wouldn't want ten thousand GPUs of its own.

178
00:23:28.333 --> 00:23:37.000
<v Sam>If a vendor told me the model's a tool and then built one, I'd ask what else they're hedging on, because I don't renew on trust.

179
00:23:37.000 --> 00:23:50.019
<v Alex>I'd read it as insurance. A router that depends entirely on rivals' models gets squeezed if those rivals raise prices or restrict access. Owning some capability, especially the small models that run on your device, reduces that risk.

180
00:23:50.019 --> 00:23:50.362
<v Sam>Hedging.

181
00:23:50.362 --> 00:24:13.317
<v Alex>It's what I do with my own vendors. I don't want one contract I can't walk away from. And our Apple episode, number thirty, argued compute was splitting into a costly cloud frontier and a nearly free tier on your own hardware. That July episode said most production queries never need a frontier model, and that the obstacle to running locally was the harness, not the model.

182
00:24:13.317 --> 00:24:17.086
<v Sam>And Perplexity's 2026 is close to a direct test of both.

183
00:24:17.086 --> 00:24:24.623
<v Alex>It built the harness, shipped it on consumer machines, and kept the frontier in the cloud for the steps that need it.

184
00:24:24.623 --> 00:24:25.994
<v Sam>So we were right.

185
00:24:25.994 --> 00:24:29.077
<v Alex>Early, and partly right. My board would take that.

186
00:24:29.077 --> 00:24:31.818
<v Sam>Right, so Perplexity does this. Does anyone else?

187
00:24:31.818 --> 00:24:47.578
<v Alex>Everyone does, and that's the point. OpenAI made routing the default experience of ChatGPT with GPT-5, in August 2025. Its system card describes a fast model, a deeper reasoning model, and a real-time router that picks one based on the conversation's type, complexity and tool needs.

188
00:24:47.578 --> 00:25:04.709
<v Sam>And every model it picks between is OpenAI's. Hold on, though. A router that sends me to the fast model is making a pricing decision as much as a quality one. Who tells me the cheap model was good enough? I've been burned before by a vendor "optimising" my experience.

189
00:25:04.709 --> 00:25:28.692
<v Alex>Nobody tells you, and that's the argument for a router that isn't selling you a model. And every one runs in OpenAI's cloud. Apple routes across the device line. Tim Cook told investors in January the rebuilt Siri will keep running on the device and in Private Cloud Compute, Apple's own servers, and the personalised Siri is a collaboration with Google, whose Gemini models sit under the new foundation models.

190
00:25:28.692 --> 00:25:39.655
<v Sam>So Apple decides what stays on the iPhone and what goes up. And as the person who has to explain where a customer's data went, that's the part I'd want in writing.

191
00:25:39.655 --> 00:26:06.036
<v Alex>That's fair, and the answer is a written policy, which is what Apple's setup is not. The models and the servers belong to Apple and one chosen partner. Google does it for developers on Android. In April it announced an experimental hybrid inference API in Firebase that routes requests between Gemini Nano on the phone and Gemini in the cloud. The first version uses simple rules. Prefer the device, fall back to the cloud, or the reverse.

192
00:26:06.036 --> 00:26:17.000
<v Sam>So rules, not a scheduler. Which I'd have bet on, because anyone shipping fast ships the dumb version first, and I respect that more than a clever one that arrives next year.

193
00:26:17.000 --> 00:26:28.459
<v Alex>Rules, and rules are easy to audit, which my head of security would like. Google says smarter routing is planned. Microsoft's Foundry on Windows gives developers local and cloud models side by side.

194
00:26:28.459 --> 00:26:34.710
<v Sam>Everyone's building the bouncer. I'd just like to know which of them lets me pick my own club.

195
00:26:34.710 --> 00:26:44.780
<v Alex>You don't, in any of them. That's the shape. Each one routes inside a stack it owns, with its own models, its own data centres, its own operating system.

196
00:26:44.780 --> 00:26:50.684
<v Sam>And Perplexity? Because so far you've described four companies doing the same thing in four walled gardens.

197
00:26:50.684 --> 00:27:10.477
<v Alex>Perplexity's the one routing across rivals. More than fifteen frontier models from competing labs in the cloud, open models on the device, Apple and Nvidia hardware today. And on phones, the Galaxy S26 ships with a "Hey Plex" wake word, next to Google's and Samsung's own assistants, and Perplexity says it powers search and reasoning for Bixby.

198
00:27:10.477 --> 00:27:23.325
<v Sam>"Hey Plex." I'm not saying it out loud in a supermarket. And three assistants fighting over one wake word is how a customer ends up arguing with a phone. Who wins that, the maker or the user?

199
00:27:23.325 --> 00:27:39.299
<v Alex>The maker, usually, which is a problem for you and a feature for them. So neutrality is the edge. A company that routes across every lab can always pick the best or cheapest model for a step. A platform owner has reasons to prefer its own.

200
00:27:39.299 --> 00:27:52.147
<v Sam>That's the piece I'd hold onto. If I'm the buyer, I need a router whose incentive isn't to push me toward its own model. Every vendor I've ever had gets less friendly the moment I can't leave.

201
00:27:52.147 --> 00:28:08.121
<v Alex>And it's also the exposure. Perplexity's router lives inside operating systems owned by Apple, Google and Microsoft, and each of them can build a switch at the system level, with deeper access to your files, your apps and your hardware than any app will ever have.

202
00:28:08.121 --> 00:28:10.204
<v Sam>Has anyone actually done that yet?

203
00:28:10.204 --> 00:28:11.941
<v Alex>Apple already has, for Siri.

204
00:28:11.941 --> 00:28:18.886
<v Sam>So you've got a guest in somebody else's house who's built the best furniture, and the landlord's started making furniture.

205
00:28:18.886 --> 00:28:21.316
<v Alex>That is my entire case, thank you.

206
00:28:21.316 --> 00:28:25.831
<v Sam>It's not your case. It's a metaphor with a case-shaped hole in it.

207
00:28:25.831 --> 00:28:27.220
<v Alex>It's a good metaphor.

208
00:28:27.220 --> 00:28:43.888
<v Sam>It's a fine metaphor. Here's mine, though. The landlord can build furniture for the landlord's own rooms. A router that has to route to its own rival has to be good at being a router, and Alex, nobody at Apple is going to route me to Google's rival.

209
00:28:43.888 --> 00:28:45.971
<v Alex>They partnered with Google for Siri.

210
00:28:45.971 --> 00:28:55.000
<v Sam>One partner they chose. I'm not asking for one partner. I'm asking whether the switch will send my customers to whichever model is cheapest this month.

211
00:28:55.000 --> 00:29:00.682
<v Alex>That's the neutral's edge, I give you that. It's a position the platform owners can contest but not easily copy without routing to their own rivals.

212
00:29:00.682 --> 00:29:11.610
<v Sam>There. We agree on something. Write it down. That's the part I'd put in a vendor brief: tell me who you route to, and tell me who you refuse to route to, and why. A neutral router that quietly favours one lab is just a landlord in a nicer coat.

213
00:29:11.610 --> 00:29:12.922
<v Alex>I'll put it on the record.

214
00:29:12.922 --> 00:29:13.140
<v Sam>Don't.

215
00:29:13.140 --> 00:29:15.107
<v Sam>Go on, scare me. What could break the template?

216
00:29:15.107 --> 00:29:29.313
<v Alex>Three things, and none of them is hypothetical. First is the router itself. A switch that decides what leaves the device has to be right nearly every time. Coverage of the Computex launch called classification the risk Perplexity has to manage. A wrong call could send sensitive material off the machine, or hand a hard task to a local model too weak to do it.

217
00:29:29.313 --> 00:29:31.062
<v Sam>And do they publish how often it's wrong?

218
00:29:31.062 --> 00:29:40.241
<v Alex>No. Perplexity's answer is layered: the open classifier, a gate that can mask or refuse, and a request for consent. And putting the classifier out lets outsiders test it. But the company hasn't published real-world error rates for the routing decision itself.

219
00:29:40.241 --> 00:29:41.990
<v Sam>That's the number I'd want in the contract.

220
00:29:41.990 --> 00:29:45.049
<v Alex>That's the number she'll want before anything goes near customer data, and she's right.

221
00:29:45.049 --> 00:29:45.268
<v Sam>Second?

222
00:29:45.268 --> 00:29:59.474
<v Alex>The hardware bar. Twenty-four gigabytes of memory for hybrid compute, a hundred and twenty-eight on the desktop box. Most laptops and phones in use fall short, so for now the local half of the router serves a well-equipped minority of professionals. Perplexity is betting the bar falls on its own. Its Portable Computer post said every chip cycle and every model release pushes this further.

223
00:29:59.474 --> 00:30:03.845
<v Sam>That bet I'd take. My copywriters' laptops are older than some of my interns, and they've survived everything so far.

224
00:30:03.845 --> 00:30:08.654
<v Alex>There's a thing there I'd push on, though, because I've lost money on that assumption. Cheaper arrives later than the slide says.

225
00:30:08.654 --> 00:30:10.184
<v Sam>You lost money on a hardware prediction?

226
00:30:10.184 --> 00:30:13.899
<v Alex>I lost two years on one, and an apology to my CFO. We can leave it there.

227
00:30:13.899 --> 00:30:14.555
<v Sam>No, we can't.

228
00:30:14.555 --> 00:30:14.992
<v Alex>We can.

229
00:30:14.992 --> 00:30:30.728
<v Alex>The third risk is legal and commercial. Perplexity has fought a run of suits from publishers and retailers. One count had nine organisations with active suits against it by the end of May. But it won an important round in August, when the Ninth Circuit vacated an injunction Amazon had won against its Comet browser agent. The court held that the user, not the agent's maker, is the party accessing the site.

230
00:30:30.728 --> 00:30:41.000
<v Sam>Which matters for everybody building an agent, not just them. My brand is going to have agents shopping its site, and the idea that the user is the party accessing it, not the agent's maker, is what keeps my legal team calm and my pricing page honest.

231
00:30:41.000 --> 00:30:52.951
<v Alex>Hugely. And the case goes on in the district court. The last piece is the layer under everything. Every frontier step Perplexity routes is bought from a lab that competes with it for the same customers.

232
00:30:52.951 --> 00:30:58.263
<v Sam>So the supplier is also the rival. I know that relationship. It's called my agency contract.

233
00:30:58.263 --> 00:31:02.910
<v Alex>Which is how you get a company with ten thousand GPUs and a hedge.

234
00:31:02.910 --> 00:31:10.878
<v Sam>So, what's the summary? That's three risks and I've lost track of the order, and I'd like to take it back to my team.

235
00:31:10.878 --> 00:31:16.521
<v Alex>One: accuracy nobody's measured in public. Two: hardware that's still arriving. Three: suppliers who are also rivals.

236
00:31:16.521 --> 00:31:22.829
<v Sam>And for my team, what does that do on Monday? Name one thing I can say in a stand-up.

237
00:31:22.829 --> 00:31:30.132
<v Alex>It means you try it on the work that can go wrong without hurting anyone. Not the quarterly numbers. The dull stuff.

238
00:31:30.132 --> 00:31:35.776
<v Sam>The dull stuff my juniors do. Which is, I'd point out, most of what my juniors do.

239
00:31:35.776 --> 00:31:38.100
<v Alex>That's the part that scares me too.

240
00:31:38.100 --> 00:31:46.731
<v Sam>My junior copywriters are the dull stuff. I know exactly who does the first draft of an autoship email, and I'd like her to stay employed.

241
00:31:46.731 --> 00:31:48.059
<v Alex>Better stake than mine.

242
00:31:48.059 --> 00:31:51.047
<v Sam>Don't undersell it. A budget line is somebody's job.

243
00:31:51.047 --> 00:32:10.965
<v Alex>So where does it point? My stores run on phones and laptops I can control, so the control point is the device. The model becomes a part I buy from many labs, and whatever assigns the work becomes the prize. Unless routing stays an open layer that any assistant can plug into, the advantage goes to whoever owns the device.

244
00:32:10.965 --> 00:32:16.277
<v Sam>And that's your case in one sentence. I'll say it's clear. I won't say it's comfortable.

245
00:32:16.277 --> 00:32:19.265
<v Alex>It's the sentence I'd have written before the show.

246
00:32:19.265 --> 00:32:36.859
<v Sam>Here's what worries me about it. Where my customers start. If the first thing they open is an assistant on their phone, not our site, then whoever owns that assistant owns my storefront. That's why a neutral layer matters to me. It's my only shot at the assistant not belonging to one company.

247
00:32:36.859 --> 00:32:42.171
<v Alex>And my answer is that the phone maker is who gets to decide whether that happens.

248
00:32:42.171 --> 00:32:49.807
<v Sam>Which is the whole fight, in one sentence. So convince me what it means beyond the tech. The economy, the politics, the labs.

249
00:32:49.807 --> 00:33:09.061
<v Alex>Economics first. The industry's record data-centre spending assumes intelligence is consumed in the cloud. Every task moved to a laptop is demand that never reaches a data centre, and my cloud bill is the thing that would notice. It doesn't make the cloud redundant, since frontier steps still need it, but it caps how much the cloud carries.

250
00:33:09.061 --> 00:33:24.000
<v Sam>Which is how you end up with Nvidia and Perplexity in the same sentence. And if the work moves onto laptops, then the people who sell laptops and chips quietly win, while the people who sign cloud contracts, like me, get a better negotiating position.

251
00:33:24.000 --> 00:33:39.330
<v Alex>The Information reported in August that Nvidia, which sells the DGX Spark that Portable Computer runs on, was in talks to invest in Perplexity at a valuation above thirty billion dollars, up from twenty billion in its last closed round. No deal has been announced.

252
00:33:39.330 --> 00:33:40.352
<v Sam>No deal. Noted.

253
00:33:40.352 --> 00:33:48.187
<v Alex>And politics. A Perplexity spokesperson said at Computex, "Sensitive and sovereign work can stay local, which changes the need for massive country-level infrastructure."

254
00:33:48.187 --> 00:33:55.341
<v Sam>That's the claim I'd want tested in public. Every town that's been promised a data centre was told it was needed.

255
00:33:55.341 --> 00:34:00.110
<v Alex>If sensitive steps stay local, governments might need less of it. For some workloads.

256
00:34:00.110 --> 00:34:02.495
<v Sam>Some. That word belongs on the brochure.

257
00:34:02.495 --> 00:34:08.286
<v Alex>And it hands regulators a concrete object to inspect. The rule that decides what leaves the device.

258
00:34:08.286 --> 00:34:24.297
<v Sam>That's my audit again. And I'd add the unglamorous version. A rule a regulator can read is a rule a company can be held to. A rule that lives in a black box is a promise, and I've signed enough promises from vendors to know the difference.

259
00:34:24.297 --> 00:34:27.022
<v Alex>That's your audit again, and I'll take it.

260
00:34:27.022 --> 00:34:35.198
<v Sam>And the labs? If I were one, I'd be furious. You spend a fortune building the thing and the customer never learns your name.

261
00:34:35.198 --> 00:34:55.297
<v Alex>If the router picks the model for each step, a lab's customer is increasingly the router, not the person. Labs compete on price and on narrow strengths. Coding here, long context there. It's the commodity dynamic from the bubble episode, and it's why OpenAI, Google and Apple each prefer a router that keeps the choice inside their own stack.

262
00:34:55.297 --> 00:34:59.725
<v Sam>So how do we know who's right? Name things I can look at.

263
00:34:59.725 --> 00:35:19.824
<v Alex>Four. One, whether Perplexity's router starts choosing on cost and speed without you picking the models. That's what makes token value per watt real. Two and three, whether the operating system owners open system-level routing to rival models or keep it closed. And four, whether open models keep closing on the frontier, which makes a neutral router more valuable.

264
00:35:19.824 --> 00:35:33.110
<v Sam>The first one's the one I like, because it doesn't need a lawyer or a conference. If a release picks the cheaper model for me and my bill drops, I'll believe in the router. Until then it's a bouncer.

265
00:35:33.110 --> 00:35:36.857
<v Alex>Next Computer releases. The OS owners' one, next year's developer conferences.

266
00:35:36.857 --> 00:35:49.121
<v Sam>So we'll know in a year, which is about how long it takes my team to finish one planning cycle. I'd rather know sooner, but I'll take a year with a clear thing to look for.

267
00:35:49.121 --> 00:35:51.165
<v Alex>More like the next two years.

268
00:35:51.165 --> 00:35:53.890
<v Sam>That's not a prediction, that's a fortune cookie.

269
00:35:53.890 --> 00:35:59.000
<v Sam>Okay. Verdict time. You first, because you've been polite for an hour and it's unnatural.

270
00:35:59.000 --> 00:36:19.981
<v Alex>I'm still on the owner of the device. I think the switch is worth most to whoever controls the place it's installed. But you moved something. The thing about enterprise administrators setting rules and auditing what leaves a device is the single thing the platform owners haven't shown me. If my CISO gets that from a neutral layer and not from the phone, I'd take it.

271
00:36:19.981 --> 00:36:45.095
<v Sam>And I moved too. I came in saying the neutral layer wins because it's on my side of the table. I'm still there on price, and on lock-in, and on the fact that I've never once been glad a vendor was the only option. But you've got me on the landlord point. If the landlord decides what the tenant can see, a neutral layer that lives inside somebody else's operating system is only neutral until they say it isn't.

272
00:36:45.095 --> 00:36:46.048
<v Alex>So we split.

273
00:36:46.048 --> 00:36:50.499
<v Sam>We split. I'll use the neutral one until my phone tells me to stop.

274
00:36:50.499 --> 00:36:52.724
<v Alex>And I'll watch whether it's allowed to.

275
00:36:52.724 --> 00:36:54.949
<v Sam>That's a very you way to end.

276
00:36:54.949 --> 00:37:11.480
<v Alex>One thing to do on Monday. Find out which of your steps are dull and cheap and private, because that's the work a router will pull off your cloud bill first. The thing I'm taking from this is the question I'll ask every vendor now: who decides which model touches my data?

277
00:37:11.480 --> 00:37:29.600
<v Sam>And mine is simpler. Whose switch is it, and can I leave it? That's what I'll ask my vendors on Monday. And it's what I'll tell my junior copywriters to ask about the agents they're about to work beside, because the people who do the dull work are the first to notice when the switch is wrong.

278
00:37:29.600 --> 00:37:45.177
<v Alex>So, the recap. Perplexity never built a frontier model. It built the switch that decides which model, and now which machine, does each step. In under five months it went from a stage demo to three products and a set of open pieces, while its revenue run rate quadrupled.

279
00:37:45.177 --> 00:38:00.118
<v Sam>It's right that value moves to whoever decides which part does which job. It's early because today's router decides mostly on privacy, and the cost-and-speed part is still a promise. And it's contested because Apple, Google, Microsoft and OpenAI are all building one inside their own walls.

280
00:38:00.118 --> 00:38:18.874
<v Alex>And the whole thing hangs on one question. If routing stays a neutral layer any device can run and any lab can supply, Perplexity's well placed to be its default. If it becomes a feature of the operating system, the switch belongs to the platform owners, and Perplexity will have shown the industry the way without keeping the prize.

281
00:38:18.874 --> 00:38:30.000
<v Sam>Thank you so much for listening. We hope you came away seeing a bit more clearly where all of this is heading. It's a fast, complicated picture, and that's what makes it worth following closely.

282
00:38:30.000 --> 00:38:47.147
<v Alex>And one note on how this is made. It's AI-generated, and so are we. Dan builds a custom stack of AI tools to research, analyse, verify and illustrate the questions worth understanding, mostly to learn them himself, and publishes it for anyone who'd like to follow along. AI-assisted, fact-checked, worth a second look.

283
00:38:47.147 --> 00:39:11.412
<v Sam>Before we go, one useful thing you can do: follow the show. Whatever app you're listening in, there's a follow or a plus button. One tap, free, and you get each new episode the moment it lands. For a small independent show, a follow is the biggest single thing that helps it reach other people trying to make sense of all this. So if this has been worth your time, go ahead and hit follow.

284
00:39:11.412 --> 00:39:24.029
<v Alex>And one last thing. If someone in your life keeps asking where AI is heading, sending them this episode is the kindest thing you can do, for them and for the show. A small show lives on every share.

285
00:39:24.029 --> 00:39:25.000
<v Sam>Until next time.
