Benedict Evans discusses AI leverage points, distinguishing transformative tech from incremental tools and highlighting prediction challenges.
Summary
Benedict Evans argues that investors must distinguish between technologies that fundamentally alter business models and those that merely provide incremental efficiency gains. He uses examples like Uber versus Airbnb and the internet's impact on newspapers versus heavy machinery to illustrate the concept of 'point of leverage.' Evans suggests that while LLMs can automate digital tasks at scale, their physical limits and long-term transformative potential remain uncertain compared to known hardware constraints. He warns against focusing on specific 'killer apps,' citing the historical misprediction of 3G use cases, and emphasizes the need for continuous reassessment of competitive advantages in a rapidly changing landscape.
Key takeaways
- 1 Where Uber was transformative to taxis, Airbnb was entirely incremental to hotels.
Benedict Evans · 7:01
- 2 AI agents act like 'million interns' that can test every possible combination of inputs to find a path, scaling problem-solving through volume rather than novel creativity.
Benedict Evans · 8:46
Episode sections
- 0:00 – 2:51
Introduction and AI as a Digital Automation Tool
Speakers introduce the podcast and frame AI discussion around the internet in 1996/1997. Benedict Evans claims LLMs can automate any digital task where responses can be validated at scale, distinguishing this from physical/digital hybrid tasks.
- 2:51 – 7:01
The Point of Leverage: Tool vs. Fundamental Change
Evans introduces the concept of 'point of leverage,' asking if a new technology gets at the core of a business or is merely a useful tool. He contrasts transformative impacts (newspapers, music) with incremental ones (heavy machinery, law firms using databases).
- 7:01 – 8:08
Case Studies: Uber vs. Airbnb and Insurance Tech
Evans compares Uber's transformative impact on taxis to Airbnb's incremental impact on hotels. He notes that while Uber changed the taxi model, hotels remain largely unchanged in their core operations (owning planes, buying fuel), similar to how insurance tech has not fundamentally altered the insurance business.
- 8:08 – 12:43
LLMs as 'Million Interns' and the Problem of Creativity
Evans describes LLMs as providing average answers or scaling 'interns' to test combinations. He questions whether AI can generate truly novel ideas outside training data, citing AlphaGo's ability to invent new moves as a rare example of validation for novelty.
- 12:43 – 14:31
Historical Context: 3G and the Wrong Questions
Evans recalls analysts asking for the 'killer app' for 3G in 2000, predicting video calling or WAP. He argues the real value was simply having the internet in your pocket, a realization that took 10 years and differed from initial forecasts.
- 14:31 – 15:42
Physical Limits and Future Uncertainty
Evans contrasts AI with mobile/smartphones, noting that while we knew physical limits for hardware (e.g., battery life, folding screens), we do not know the physical limits of AI models in 2-5 years, making prediction difficult.
- 15:42 – 17:04
Custom Tools and Competitive Advantage
Evans discusses why large law firms might build their own AI tools: to avoid using the same generic models as competitors. He draws a parallel to investment banks using unique valuation models, suggesting that custom AI could be a point of leverage.
- 17:04 – 17:55
Conclusion: The Right Questions for Investment
Speakers conclude that the key is identifying which questions matter and which are 'wrong' or irrelevant. Evans reiterates that current frameworks may become obsolete quickly, requiring continuous reassessment of leverage points.
Claims worth verifying
LLMs can automate anything digital where you have enough training data and can validate the response at scale.
Benedict Evans ·1:52
How to verify: Test LLM capabilities on specific digital tasks (e.g., coding, copywriting) to see if outputs can be mechanically validated against ground truth at scale. For recent events, verify 2026-08-10 through 2026-11-10. Use another period only when the speaker explicitly names it.
The internet was transformative for the music and newspaper businesses but merely a useful tool for heavy machinery manufacturers like Zeemans.
Benedict Evans ·3:18
How to verify: Review historical revenue data for newspaper and music companies post-internet adoption vs. heavy machinery manufacturers; check Zeemans' business model for internet dependency. For recent events, verify 2026-08-10 through 2026-11-10. Use another period only when the speaker explicitly names it.
Revolut is a giant business but has not caused traditional banks to disappear.
Unidentified speaker (UNKNOWN) ·4:26
How to verify: Check Revolut's user base and revenue growth against total banking sector deposits and profitability; look for bank closures or mergers directly attributed to Revolut. For recent events, verify 2026-08-10 through 2026-11-10. Use another period only when the speaker explicitly names it.
Uber was transformative to taxis, while Airbnb was incremental to hotels.
Benedict Evans ·7:01
How to verify: Compare market share shifts in the taxi vs. hotel industries post-platform adoption; analyze hotel revenue models for disruption vs. taxi model changes. For recent events, verify 2026-08-10 through 2026-11-10. Use another period only when the speaker explicitly names it.
AlphaGo could invent a new move and test it using an internal scoring system.
Unidentified speaker (UNKNOWN) ·10:12
How to verify: Review AlphaGo research papers (e.g., Nature publications) detailing its self-play training and move generation capabilities. For recent events, verify 2026-08-10 through 2026-11-10. Use another period only when the speaker explicitly names it.
In 2000, analysts predicted video calling and WAP as killer apps for 3G.
Benedict Evans ·13:04
How to verify: Search for IDC or Frost & Sullivan reports from 2026-2026 predicting 3G use cases; compare with actual 3G adoption data. For recent events, verify 2026-08-10 through 2026-11-10. Use another period only when the speaker explicitly names it.
Meta launched Muse as a consumer agentic platform.
Unidentified speaker (UNKNOWN) ·15:05
How to verify: Check Meta's official announcements or press releases for a product named 'Muse' described as a consumer agentic platform. For recent events, verify 2026-08-10 through 2026-11-10. Use another period only when the speaker explicitly names it.
Large law firms are building their own AI tools to avoid using the same models as competitors.
Unidentified speaker (UNKNOWN) ·15:57
How to verify: Review press releases or job postings from major law firms (e.g., Freshfields, Linklaters) regarding proprietary AI tool development. For recent events, verify 2026-08-10 through 2026-11-10. Use another period only when the speaker explicitly names it.
European mobile operators spent 110 billion euros on 3G licenses.
Tony Karen Brown ·12:43
How to verify: Check European Commission or national regulatory body records for 3G spectrum auction revenues in the early 2000s. For recent events, verify 2026-08-10 through 2026-11-10. Use another period only when the speaker explicitly names it.
LLMs provide average answers or act as 'million interns' rather than generating truly novel ideas outside training data.
Unidentified speaker (UNKNOWN) ·8:08
How to verify: Conduct A/B testing of LLM outputs for creative tasks against human-generated novel content; analyze training data overlap with outputs. For recent events, verify 2026-08-10 through 2026-11-10. Use another period only when the speaker explicitly names it.
Do your own research
- Review Benedict Evans' past writings or interviews on 'point of leverage' to understand his framework for tech adoption curves.
- Compare Uber's and Airbnb's financial filings to assess their actual impact on incumbent profitability versus market share.
- Analyze historical analyst reports from 2000 regarding 3G predictions to verify the accuracy of 'killer app' forecasting.
- Investigate current AI model training data requirements and validation mechanisms for specific digital tasks mentioned.
- Check recent filings from major law firms or investment banks to see if they are building proprietary AI tools versus using generic models.
- Research the physical constraints of current AI hardware (e.g., battery life, cooling) to benchmark against Evans' claims about unknown limits.
- Look for case studies of companies that failed to adapt to the internet despite having digital presence, such as heavy machinery manufacturers.
- Verify the performance metrics of AlphaGo's novel moves against traditional Go strategies to understand the validation mechanism Evans cites.
Full transcriptexpand
- 0:00
- Tony Karen Brown: Hi, I'm Tony Karen Brown.
- 0:02
- Benedict Evans: And I'm Benedict Evans.
- 0:04
- Unidentified speaker (UNKNOWN): And
- 0:04
- Tony Karen Brown: we're going to talk about AI.
- 0:05
- Benedict Evans: Uh what else? Yes. Wet labs, Doom, giant one-eyed aliens. Um
- 0:12
- Tony Karen Brown: the end of the world and civilization as we know it.
- 0:15
- Benedict Evans: I sp I spend a lot of time talking to big companies about AI or listening to them talk about AI. And there's the this sort of, okay, we're doing this stuff now. and next year, but what's the fundamental change conversation? Which is very like, as I say, in a lot of contexts, it's like looking
- 0:35
- Benedict Evans: at the internet in 1996 or 1997, which is kind of where we are with AI. It's a very big deal. We don't really know how all of this is going to work and what the building blocks are going to be, nor what it will mean for particular industries. There's an old joke that the newspapers thought the internet was going to be great because they're printing. their bills would go down. And the challenge there is, you know, you
- 0:55
- Benedict Evans: can say very vague general things like, you know, distance will mean less and mod physical assets won't matter in the same way, which is like kind of true but not very helpful. Or you can try and make super specific predictions about individual industries and, you know, do a big like consulting exercise and make big matrices and like you'll get it all wrong. Um
- 1:15
- Benedict Evans: and but there's sort of something in between which is to think, well, let's go push a little bit further on what are the kind of kind of conceptual building blocks to think about what's exposed and not exposed.
- 1:24
- Tony Karen Brown: Yeah. And it feels like every week we have like a good couple of questions that sit at that weird middle ground
- 1:29
- Unidentified speaker (UNKNOWN): intersection.
- 1:30
- Benedict Evans: Yeah. And
- 1:31
- Unidentified speaker (UNKNOWN): we've got a few
- 1:31
- Tony Karen Brown: this week.
- 1:32
- Benedict Evans: Well, so there's a sort of a high-level point, which is it looks like LLMs are going to be able to do everything. anything, at a minimum, anything digital where you have enough training data and can validate the response at scale. Maybe that's a um duplication. If it's digital, you probably can if it's not digital, it gets much harder to
- 1:52
- Benedict Evans: validate the response at scale. But if it's digital, if you can capture it, if it's explicit Implicit and not implicit, if you can test the results at scale mechanistically, you can probably almost certainly automate this with LLMs. And so that's one point. That's a kind of a one fairly high-level way to think about it.
- 2:12
- Benedict Evans: And you push that one step further, and you can say, well, we're going to have much faster, much cheaper, and automation of many more things. And just at a kind of base level, that might change a lot of. industries. And one can kind of dig into what that means and the Jevons paradox and price
- 2:31
- Benedict Evans: elasticity and so on. But it struck me that there's a sort of there are two further questions to ask. And one of them is a question that applied to the internet in particular as well and probably PCs, which is you can make and we've talked about this in a different context in the past, you can make an AI-enabled law firm or you
- 2:51
- Benedict Evans: might have created an internet-enabled law firm. Is the AI or internet or the PC getting at the key point of leverage in your business?
- 3:00
- Benedict Evans: And clearly there are some industries where the internet unlocked something fundamental, maybe not obvious to begin with, but it unlocked Unlock something fundamental that created completely new possibilities, new market, new competition, demolished your model. And there are others where it didn't. So
- 3:18
- Benedict Evans: for the sake of argument, so
- 3:20
- Tony Karen Brown: you mean like the services and the product completely changed the business versus
- 3:24
- Benedict Evans: Does this new technology actually get at the core of your business and unlock something? Maybe what you thought was the core, maybe something else, and unlock something crucial. Or
- 3:33
- Unidentified speaker (UNKNOWN): is
- 3:33
- Tony Karen Brown: it additional to tool?
- 3:34
- Unidentified speaker (UNKNOWN): the other? And the
- 3:34
- Benedict Evans: answer is, and the answer is sometimes it isn't. So if you get go back and think about the internet, um you could say that obviously the internet was transformative to the music business and the newspaper business and so on. It was but the other extreme, if you made heavy machinery, if you made construction equipment, if you were Zeemans and you made gas turbines, this is kind of a useful tool. Everyone you get email now, email's great, but it
- 3:54
- Benedict Evans: doesn't actually change the business of making 20 or 30 gas turbines a year and selling them to power utilities. And then there are people who are kind of in the middle. So you could think about, say, banks or airlines, where the internet unlocks neo banks, and it means you can make a new consumer banking business without needing lots of physical branches.
- 4:14
- Benedict Evans: Or now you have, if you're an airline or a hotel, now you have these new aggregators and new routes to market and new people who want commission from you and new ways to sell and get customers.
- 4:26
- Unidentified speaker (UNKNOWN): But are the
- 4:26
- Benedict Evans: at the end of the day, your business is still owning airplanes and landing slots and buying fuel and filling of seats. And yes, Revolut's a giant business, but like my parents haven't switched from their bank and you know the other banks are still there. So it was an entry path.
- 4:41
- Tony Karen Brown: It's a similar business for a new audience that's looking for more online digital banking first and only and take it from there.
- 4:47
- Benedict Evans: It's marginal change. It may be a new entry point it may
- 4:51
- Unidentified speaker (UNKNOWN): be
- 4:51
- Benedict Evans: a new route to market, it may be a new competition, but it doesn't actually completely overturn everybody else. It doesn't mean that all the incumbents go away in the way that we saw with say with newspapers or record companies.
- 5:01
- Tony Karen Brown: It's a new way of doing the same business. Yeah.
- 5:03
- Benedict Evans: Yeah. So so you might be John Deere, you might be a bank, you might be a newspaper. And I think that's a kind of a a way of obviously it's never as simple as that. And as a consultant everything has to be three bullet points, but that's a kind of a way of asking, like just because just saying the internet may or may not actually change everything. And probably making an internet enabled law firm was a great way of getting
- 5:23
- Benedict Evans: customers, but
- 5:24
- Unidentified speaker (UNKNOWN): at
- 5:25
- Benedict Evans: the end of the day, fresh fields and link latest are still there. You know, the big law firms are still there. Um and so this is the sort of the first question is well, okay, how does this change everything that what happens if you're doing everything faster and cheaper? The second is does that actually unlock some fundamental point of leverage in your
- 5:44
- Benedict Evans: business? Does it actually change the thing that matters? Um, I had this conversation on LinkedIn the other day with a lawyer because I was sort of wondering what did legal databases do to law firms? Because, you know, in 1980, you want a precedent, you send an associate to the law library for a day or two.
- 6:03
- Benedict Evans: And then in 1990 or 2000, you have a database and you're paying for the database, but you know, you can get the precedent much quicker. And the answer is actually that was a really small part of the business, so who cares?
- 6:14
- Unidentified speaker (UNKNOWN): I don't
- 6:14
- Benedict Evans: know if that's true, but that was the answer. But it might be. That's kind of the point. Maybe the internet, the thing that the internet changes isn't important. And this may be Maybe the case for AI as well for some industries. Like this doesn't change anything that matters. Or it changes everything, but it doesn't change. the like it doesn't create new entrants. It's just like we like you could say that computing changed everything about accounting, except it's all the same
- 6:34
- Benedict Evans: companies and all the same people billing all the same customers the same amount of money.
- 6:37
- Tony Karen Brown: So you're saying sort of like in some cases it's a useful tool, but
- 6:41
- Benedict Evans: some cases, yeah.
- 6:42
- Unidentified speaker (UNKNOWN): And
- 6:42
- Tony Karen Brown: then some of the product has shifted.
- 6:44
- Benedict Evans: Exactly. And then the third qu actually I I was at an event yesterday for an in uh shirt people who make a software for the insurance industry.
- 6:52
- Unidentified speaker (UNKNOWN): And
- 6:52
- Benedict Evans: I gave a version of my new AI presentation and one of the first questions was like we we all heard a bunch about machine learning ten years ago and that didn't happen.
- 6:59
- Tony Karen Brown: So what's changed?
- 7:01
- Benedict Evans: Why did insure tech, why did machine learning not completely change the nature of the insurance business? Um there's another c sort of chart I somet slide I sometimes use, which is to compare the impact of Uber on taxis with the impact of Airbnb on hotels. Where Uber was transformative to taxis, Airbnb was entirely incremental to hotels. And there's a whole bunch of
- 7:20
- Benedict Evans: reasons why that was, starting with this insurance event yesterday was in a hotel. And it wasn't in an Airbnb.
- 7:27
- Tony Karen Brown: Oh, that's a good
- 7:28
- Unidentified speaker (UNKNOWN): point.
- 7:29
- Tony Karen Brown: That's such a good example.
- 7:31
- Benedict Evans: And you know, when you travel or when my wife travels for business, you arrive at you know eight o'clock at night and you want a gym and room service and security, and you're gonna leave the client side at six o'clock the next morning. No, you're not staying in an Airbnb in the suburbs of Charlotte. Um so it depends. I think the third question is, and it
- 7:50
- Benedict Evans: says that's that's a point that th those are both points that sort of apply to every new technology in one way or another. I think the third question is much more specific and conceptual about the nature of machine learning and and LLM. LLMs, which is these things are telling us what most people would probably say.
- 8:08
- Unidentified speaker (UNKNOWN): And
- 8:09
- Benedict Evans: sometimes
- 8:10
- Unidentified speaker (UNKNOWN): you mean it's
- 8:11
- Tony Karen Brown: giving us an average of how people think and what
- 8:13
- Benedict Evans: it's this is what a good answer would probably look like. Sometimes that's what you want. If you're doing your taxes, you want what a CPA would probably tell you. Um, sometimes it's what would a million people probably say. This I was used to describe the
- 8:26
- Unidentified speaker (UNKNOWN): last
- 8:26
- Benedict Evans: way. wave of machine learning as giving you infinite interns, like listen to every call coming into the call center, tell me if the customer's angry. 10-year-old can do that. And that's one way to think about all of the cyber stuff that's going on, which is it's not that this thing is amazingly clever, it's just there's 10,000 of them and they're really fast. You spin up
- 8:46
- Benedict Evans: 10,000 agents and they try every possible combination of every possible thing until they're they find the path. That's your million interns as million cyber interns, perhaps. Um, you could also say that's part of scaling.
- 8:59
- Tony Karen Brown: And your question here is is the desired answer outcome that average, or is it something else that you're looking for?
- 9:06
- Benedict Evans: Yes. Do we want something that isn't in the training data? And this is, you could argue, the sort of the essence of creativity is something that is different but good. And the theoretical problem there is that you can very easily imagine a model making more generic pop music
- 9:26
- Benedict Evans: or more generic images like that. Um, more difficult or certainly a very different level of abstraction would be for a model to say, hey, we've all been listening to disco and um.
- 9:39
- Benedict Evans: um progrock and abba and it's time f and the economy's turned down and it's time for something completely different. So here are the sex pistols. And again, playing more stuff that sounds like the sex pistols is is a much easier problem than inventing the sex pistols. Would it know this isn't in the training data, but people would like it. Can
- 9:59
- Benedict Evans: you build a model of taste
- 10:02
- Tony Karen Brown: based on
- 10:03
- Benedict Evans: that can evaluate something new? Or can you only evaluate how well it matches what we or the things we already know people like? Um
- 10:12
- Unidentified speaker (UNKNOWN): this
- 10:13
- Benedict Evans: is of course I often go back to thinking about AlphaGo that AlphaGo could invent a new move, but it could test it. This is the validation point. It could make a new move, a new strategy that people had never done before that people thought was a mistake, but it had an internal scoring system, so it could test. whether that was a good move what would that equivalent be for making something that would appeal
- 10:32
- Benedict Evans: to people how would we know people would like that but it doesn't match the training data there's a the other side of the brain I never remember if it's left or right there's a discussion now of like can you make new scientific discoveries or can only can you only expand based on what we already know? So I think there was a paper
- 10:52
- Benedict Evans: in the last week or two of someone sort of trying to give a model what we knew about physics before Einstein and could it make the jump to Einstein? I'm not sure whether you can conclusively test the answer to that. And of course, people didn't make the jump either until Einstein happened. But this there's a sort of theoretical question of how far can
- 11:12
- Benedict Evans: you get from just extrapolating. You can certainly get some kinds of new discoveries with a million interns. And again, this is a conversation about theoretical mathematics that was happening
- 11:22
- Unidentified speaker (UNKNOWN): last
- 11:22
- Benedict Evans: week. Um, I am totally unplaced to take a view on this, but there is a sort of question of are there some kinds of maths that LLMs are good at finding new things?
- 11:33
- Benedict Evans: Is it constrained to certain sorts of problem where the answer is just test everything? Or is it something else?
- 11:40
- Tony Karen Brown: And I wonder if we already know this. Like if we're aware of, you know, the instances and the use cases and the moments where we want and we're looking for the average or we're looking for more of the same. And what are the instances where no no we need you to go beyond the average answer, the normal answer, we're looking for something completely new and different. I wonder if that's going to shift how we approach these things.
- 12:00
- Unidentified speaker (UNKNOWN): Yes, I think you
- 12:01
- Benedict Evans: could probably propose that if you were at Anthropic and you'd asked Claude, do you think it's a good idea to start a wet lab researching human diseases at the same time as telling everybody we need to slow down AI research? I'm pretty sure Chat GPT three would have said, yeah, probably not a good idea. Seriously
- 12:19
- Benedict Evans: read the room, guys.
- 12:23
- Unidentified speaker (UNKNOWN): There are
- 12:23
- Benedict Evans: were some questions
- 12:25
- Unidentified speaker (UNKNOWN): to read the room. Yes.
- 12:26
- Tony Karen Brown: Hopefully you get to the same answer.
- 12:28
- Unidentified speaker (UNKNOWN): In fact,
- 12:28
- Benedict Evans: you probably only need only needed one intern
- 12:31
- Unidentified speaker (UNKNOWN): to answer
- 12:32
- Benedict Evans: that question. Like somebody that you hadn't already shared a group house with. Um there's but but but but but kind of zooming out, I think this is a these will be the wrong questions.
- 12:43
- Tony Karen Brown: Yeah, yeah.
- 12:44
- Unidentified speaker (UNKNOWN): I
- 12:44
- Benedict Evans: mean I always remember when I was a baby analyst looking at mobile in 2000 and the European mobile operators. had just spent 110 billion euros on 3G licenses, which at the time seemed like a lot of money. Not so much now. And every client would say, okay, well, what's the killer use case for 3G? What's the killer app for 3G? That was
- 13:04
- Benedict Evans: the question.
- 13:06
- Unidentified speaker (UNKNOWN): And
- 13:06
- Benedict Evans: is it video calling? Will people do video calling? Is it games? Is it games? Is it the web? Well Is it is it WAP? Is WAP the killer app for 3G?
- 13:15
- Benedict Evans: And the answer was that the killer app for having the internet on your po in your pocket anywhere you went was having the internet in your pocket anywhere you went. That
- 13:23
- Tony Karen Brown: was it.
- 13:23
- Benedict Evans: But it had to be done in a bunch of ways that were completely different from what anybody was thinking about then and would basically take 10 years. None of this is going to take 10 years, but
- 13:33
- Unidentified speaker (UNKNOWN): it
- 13:33
- Benedict Evans: was the wrong question.
- 13:34
- Unidentified speaker (UNKNOWN): Or
- 13:35
- Benedict Evans: it was the wrong way of thinking about what the question was.
- 13:37
- Unidentified speaker (UNKNOWN): Um
- 13:38
- Benedict Evans: go forward. five years. I've got a research paper from like IDC or Frost and Sullivan or one of those companies that's got five pages doing a forecast of how much money mobile operators would make um by um from um location lookups. Like they would charge you
- 13:53
- Unidentified speaker (UNKNOWN): ten
- 13:53
- Benedict Evans: cents every time your phone asked for a location lookup because you would be doing it from the cell.
- 13:58
- Unidentified speaker (UNKNOWN): And
- 13:58
- Benedict Evans: they said no one there wasn't like G GPS is a is a thing. There was no so so so these are all like uh we don't know, but do
- 14:09
- Tony Karen Brown: you think we fully understand? We don't know, but do you it's interesting hearing you say, look, this this was gonna it it it was gonna be the thing and it took 10 years. Do you think we fully realize that none of this is gonna take even probably a year or two? Because 'Cause I'm s hearing your conversation even with your with with your guys yesterday of just like, okay, this is where we're at today and this. is where we think we'll be in a year. We have no
- 14:29
- Tony Karen Brown: idea. It's moving way too fast.
- 14:31
- Benedict Evans: Yeah. And it's moving fast in multiple dimensions. It's moving part and of course the difference from the internet and from from mobile and smartphones is that with those technologies, as I've said before, we understood what the physical limits of the technology were. So we didn't know what the next iPhone would be,
- 14:45
- Unidentified speaker (UNKNOWN): but we knew
- 14:45
- Benedict Evans: like it wouldn't fold, it wouldn't fault. fold Because we knew that technology didn't exist yet. And we knew, you know. it wouldn't have a one-year battery life or a retinal projector or something. Whereas with this technology, we don't really know what it's going to look like in two,
- 14:59
- Unidentified speaker (UNKNOWN): three, four,
- 15:00
- Benedict Evans: five years' time, like what is physically possible for the models to be able to do.
- 15:04
- Tony Karen Brown: Yeah.
- 15:05
- Unidentified speaker (UNKNOWN): We also
- 15:06
- Benedict Evans: um, but but back to you know 1997, you know, is it browsers, is it app stores, is it this, is it that? Is it consumer video? Is it uh isn't is it an agent? Meta's launched, Muse as a consumer agentic platform. Great work. It's a lot of new habits, so asking people to form all at once. Um, and so there's there's a big
- 15:26
- Benedict Evans: difference between saying everybody's going to have this and probably quicker than last time, and And actually knowing, well, what does that actually mean? And which companies get broken apart. And this is why I kind of come back to these questions. Like,
- 15:42
- Unidentified speaker (UNKNOWN): if
- 15:42
- Benedict Evans: you're making an AI enabled law firm, I know why you're doing that, because it's really hard to change how an existing professional services company works. But is that actually going to change the fundamentals of what it is that the customer's buying from you? I mean, maybe I can argue you both sides of that,
- 15:57
- Unidentified speaker (UNKNOWN): but
- 15:58
- Benedict Evans: we don't think we know yet. And is the AI itself going to give you the going to going to give you answers that are different enough or that add enough value? And this is why you see a couple of the big law firms saying they want to make their own tools because then what they don't want is for everyone to be using the same tools that gives the same answers.
- 16:17
- Unidentified speaker (UNKNOWN): That might be
- 16:18
- Benedict Evans: a little bit like an investment bank saying we want to have our own um spreadsheet or we want to have our own model. Well we you probably do have your own valuation models, but you use the same software to build
- 16:29
- Unidentified speaker (UNKNOWN): it.
- 16:30
- Benedict Evans: And what is the
- 16:32
- Unidentified speaker (UNKNOWN): point
- 16:32
- Benedict Evans: of leverage? Do you want the average answer? So maybe the average answer in itself is a point of leverage. That is
- 16:38
- Unidentified speaker (UNKNOWN): one of the leverage points is that
- 16:40
- Benedict Evans: I actually just want it done the same way as everybody else, but I'm going to do something else. Um, and so but those are the sort of puzzles we had with the internet or with mobile. Um okay, you're using the phone, you're using mobile, you're using location. Okay, is that the key thing that matters?
- 16:57
- Unidentified speaker (UNKNOWN): Um,
- 16:58
- Benedict Evans: or is the key thing that matters somewhere else? Um and that seems like that's the same question this time around.
- 17:04
- Tony Karen Brown: And I like the way you'd framed both of like those three points that were framed of
- 17:09
- Unidentified speaker (UNKNOWN): what
- 17:09
- Tony Karen Brown: do you get and what do you not get with a million interns? And where's the question that actually matters? Um, I thought that was a nice way of framing of framing how you think about this.
- 17:18
- Benedict Evans: Yep. Until the next time.
- 17:20
- Tony Karen Brown: Well, until next week, until something minor and big changes.
- 17:23
- Benedict Evans: That's funny. I go back and look at my old presentations and sometimes and again like, well,
- 17:28
- Unidentified speaker (UNKNOWN): that
- 17:28
- Benedict Evans: was right. That was the wrong question.
- 17:30
- Tony Karen Brown: Yeah, because that's that's that's the name of the game
- 17:32
- Unidentified speaker (UNKNOWN): right now. What the Right. right
- 17:33
- Tony Karen Brown: questions to think about this week and maybe next week we have a bit more information that the questions are actually no longer valid. or relevant.
- 17:40
- Unidentified speaker (UNKNOWN): Yeah,
- 17:40
- Benedict Evans: I was like the gra I like the Groucho Marks line, these are my principles and if you don't like them I've got others.
- 17:45
- Tony Karen Brown: There you go, folks. These are questions this week. And if you don't like them, tune in next week, we'll have others.
- 17:50
- Unidentified speaker (UNKNOWN): Okay.
- 17:51
- Benedict Evans: There's a
- 17:51
- Unidentified speaker (UNKNOWN): good
- 17:51
- Benedict Evans: point to end. Good to chat.
- 17:53
- Tony Karen Brown: Good to chat. See see you soon.