The Ledger
Every statement that passed quotation and attribution checks. Mix any filter with any other: certainty 1/5, debate potential 5/5, or both at once.
why aren't all 537 resolved? a statement only gets an assessment when the public record can support or contradict it. opinions and what-ifs never can, and 0 checkable ones are still open, waiting for their date. predictions held up or didn't; assertions are supported or contradicted. on every card: ▮▮▮▮▮ certainty · ▮▮▮▮▮ debate potential. speakers are clickable
Evans: All AI questions have no answer or follow past platform shifts
“All AI questions have one of two answers. The answer is either no one knows, like how much better will the models get, for example you know, what happens to, you know, does inference cost collapse and we no longer need any more in video chips? No one knows. An…”
Evans: Tech regulation must target specific applications rather than AI monoliths
“It's my point about what do you mean by, when you say AI, what's the right level of abstraction here? Should you be talking about regulating AI, or should you be talking about, we don't want people using face recognition to decide who can get into a conference…”
Evans: LLMs generate probable answers, unlocking value only where exact correctness isn't required
“These models are very good at automating certain kinds of things, but not others. They get stuff wrong, or rather what they do is they say, what would an answer to this question probably look like? That's about the best way I can come up with formulating it ri…”
Evans: Bottom-up software sales only captures 5% of enterprise users
“There's this whole sort of thesis of bottom up enterprise software sales that you don't need to sell to the CIO. You make this amazing cool tool and the users will discover it and use it. And what actually happens is that most that that always works for like t…”
Catton-Brown: Sports copying F1's Netflix boom ignore the pandemic's role
“All of these other sports are trying to recreate this magical Netflix strive to survive moment. But they seem to omit the fact that we had a pandemic. Everyone was stuck at home. Everyone was dying for content. The first season of Drive to Survive was actually…”
Building a search index 75% as good as Google costs billions
“And so just being, like, 75% as good as Google costs hundreds of millions of dollars or billions of dollars just to have the whole thing. It's like, it's, there's a fixed cost to indexing the web.”
Evans: $1M enterprise consulting spend buys GenAI development, not deployment
“Because a million dollars gets you developing, it does not get you deploying, even from, even at Accenture's rates.”
Evans: Generative AI requires error tolerance or easily verifiable mistakes
“So you need to have a use case where either it doesn't matter if it's not a hundred percent right, which is, you know, Give me 50 ideas for slogan for a new brand of toothpaste. Like there's not, you're not, there's not like a complete. There's no right answer…”
Evans: Bottom-up software adoption caps at 5-10% TAM before needing enterprise sales
“And there's been like two things in the history of enterprise software. This has worked. I mean, one of them is Slack and I forget what the other one is. And what happens is you get to like five or 10% of your tan, and then you run out of the kind of people wh…”
Evans: Generalized EU tech rules become complex when platforms implement them
“The EU is trying to say, look, let's just have like a kind of general, generalized rules that will sort of try and capture lots of different questions at once. The problem is, of course, as soon as you get into how does Apple actually implement this in the app…”
Brown: TikTok Gives Americans a Taste of Foreign Tech Dominance
“Saying that, like, basically Americans are experiencing what the rest of Europe has experienced over the last 10 years of, yeah, your citizens are loving a piece, a social media app that wasn't designed by Americans for Americans with Americans in mind.
yeah,…”
Evans: Mandating training data disclosure will not make AI models auditable
“So it's sort of easy to say, oh, well, you have to publish what your training data is, but that's not Going to produce something that's not going to produce something that's easy for you to audit.”
Evans: AI prompt moderation will repeat social media governance battles
“You re-run all of the arguments we had around content moderation and social media of what should people be allowed to say. You either say more or less anything, which is Elon Musk's idea, or you end up with a long list of criteria, which is where Facebook has …”
Benedict Evans: Tech has not fundamentally disrupted the book publishing industry
“Tech has basically not been a problem for the book industry. You know, obviously lots has changed, and they've got Amazon as this one giant customer, but we did not move to a model in which everybody gets all the books they want, and all the books are digital,…”
Evans: Flat-rate subscriptions cap high-spending whales, failing in books
“With all of these subscription models, you're capping your whales. So with music, you had a small number of people who were paying a hundred dollars a month or more, whatever it was, and now they can only, for CDs, now they can get everything for 10 dollars a …”
Devins: Meta has the right price; Apple has the right device
“And what I wrote last summer was that Meta basically has the right price and is working towards the right device. And Apple has almost the right device and is working back towards the right price.”
Devins: 2D is the fundamentally correct medium for text and data
“The right way to present text and pictures and data is in two D.”
Evans: ChatGPT is a giant job creator for lawyers and consultancies
“ChatGPT is like a giant job creation scheme for lawyers and consultancies. Because the lawyers have to work out who we're suing, and what all the new contracts look like, and the consultancies, and people actually have to go and deploy all of this stuff.”
Evans: Regulating AI Is the Wrong Level of Abstraction
“I think regulating AI, quote unquote, is the wrong level of abstraction. To me, that's like saying, well, look, you know, FTX did all this fraud with spreadsheets, so clearly we need to regulate spreadsheets.”
Evans: Fediverse Model Fails Because Client Fragmentation Breaks Social UX
“The problem with the Fediverse model is so much of the experience design has to be in the client. And if it's distributed and there's 50 clients, that doesn't work in the same way.”
Evans: Engagement signal graphs matter more than raw content
“If you're trying to build a competitor to Instagram, I wouldn't want Tony Karen Brown's posts. What I want is who's liked Tony Karen's posts, Tony's posts, and what else did they like? And who liked their posts, and who else liked that? There's like these sort…”
Evans: Ungated, algorithm-free feeds inevitably collapse under content volume
“If you're building a recommendation system that says anybody can publish anything they like, and there's no gatekeeping, and there's no algorithms, guess what? You've got 10,000 things in your feed every day, and you can't find anything. It's like you're tryin…”
Evans: Creator reach declines due to user competition, not algorithm malice
“The other half of it is the reason that you're downranked in the newsfeed is because people don't care, A, people don't care about you that much, B, they followed 50 of your competitors, and B, they have a hundred friends. And there's no solution to that.”
Evans: Machine learning gives you 'infinite interns'
“The way I always used to talk about machine learning is it gives you infinite interns. You can get an intern to do something. Anything you can get an intern to do, machine learning can do for you know.”
Evans: AI prompts are just a return to the command line
“And so this point is a command line again.
And this is Bill Gates's point that GUIs replace command lines.
Well, you get, you've just given me a command line now.
That's what a prompt is.”
Evans: Chatbots are the wrong interface for many use cases
“I mean, there's a bunch of use cases where it doesn't matter how good the model is, a chatbot is just not the right interface.”
Brody: Sustainable products fail if not perceived as best in class
“When people are forced to buy things that they think are better for the planet or better for them, if the product isn't perceived to be best in class, nobody cares.”
Network-effect winners are threatened by unbundling, not direct clones
“Once you've won a space that has strong network effects, then the threat is not from something that's the same, but 10% better. The threat is from something that either solves the underlying problem in some completely different way or complete or unbundles you…”
Trending topics and search are the primary drivers of toxic dogpiling
“Trending topics and search are the two ways things that tend to drive dog piles. And tend to drive everyone rushing to scream at somebody.”
News on social platforms creates aggravation without driving revenue or usage
“Speaking from a business perspective, news is a lot more trouble than it's worth. It's not actually particularly good for business because it's a huge amount of aggravation and toxicity and anger and problem. And it doesn't actually drive a lot of, it certainl…”
Threads gains an advantage as ideological extremes self-select out
“All people who think that Elon Musk is brilliant, and if you point out that any particular problem or mention him in anything but less than sycophantic terms, they're all going to stay, they're not going to go to Threads. All the Nazis are going to go, I'll st…”
Benedict Evans: Vision Pro mirrors the 1984 Mac's uncertain utility and price
“In that sense, it's actually, you could kind of compare it to the Macintosh, which launches in 1984 and was over 7000 dollars in today's money. And most people still didn't know why they needed it. And I think Apple only sold like half a million in the first y…”
Benedict Evans: Productivity's future is generative AI, not bigger VR spreadsheets
“I don't think the future of productivity is seeing more columns in Excel. I think it's having an AI that builds the model for you.”
Benedict Evans: Assessing AI's commercial impact requires parking the AGI debate
“To actually kind of make any questions about what companies are going to happen and how we're going to use this, you always have to kind of park the AGI conversation, because if it happens, then all bets are off.”
Benedict Evans: Generative AI search is risky because errors are undetectable
“The challenge is as I always, I've always been saying, if you use this as a general search, then the error rates might matter and you can't tell.”
Benedict Evans: Professional AI workflows require UI controls, not prompt boxes
“As soon as you move away from make this shot better, you're probably moving away from a prompt box, and you're probably moving to buttons, and switches, and options, and palettes, and lists of possibilities, and a screen that says, okay, here are six different…”
Benedict Evans: ChatGPT's chat UI deceptively implies definitive answers
“To me, this is part of the challenge of ChatTPT versus Google is that Google says here are 10 links that might be the answer. Whereas ChatGPT says this is the answer, and it's, this is the answer. But it's not, you could argue that what it should be saying is …”
Evans: Midjourney does not threaten art because tools cannot replace artistic sensibility
“Anybody can buy a camera, but that doesn't make you a photographer. You know, anybody can fire up mid journey. That doesn't mean you're going to, you know, doesn't mean you're going to produce something great in the way that somebody who has the right sensibil…”
Evans: The nature of art is that if it can be copied, it is not art
“The nature of art is that you can't copy it. If you can copy it, it's not art.”
Cairn-Brown: Prompt engineering creativity determines generative AI output quality
“We're all being given the same tools, but you're, that creativity behind the prompt engineering is what's going to make the absolute difference here, which is fascinating.”
Evans: GPT aims for AGI by ingesting all collective human text
“The theory that all of this kind of is on a track to AGI in a sense is that you can create AGI out of the collective Knowledge of humanity, which is sort of what GPT is trying to do. It's just to give it all the texts that there is on the internet and like int…”
Evans: Generative AI generates patterns without structural understanding of underlying objects
“It's making a pattern without any structural understanding of what the actual thing actually is.”
Evans: AI utility depends on whether errors matter and can be spotted
“This is a two by two matrix, which is, you know, does the error matter and can you spot the error?”
Evans: Incumbents treat AI as a feature, but disruption requires new products
“The incumbent's always trying to make the new thing a feature. And so that's very obviously what's happening with Microsoft and Google, including this in PowerPoint and office and of course in search and everything else. But what actually happens is, is then p…”
Evans: AI summarizing news without links raises deeper issues than search engines
“If you're going to, you know, build a new system that will just tell you what the story was without linking or giving you credit, that's very different from you know, what Google and Facebook do, which is you search for a new story and they tell you where to r…”
Evans: Generative AI models cannot be deterministically controlled using crude keyword filters.
“You can't, you know, you cannot deterministically control what it says. So you can kind of put crude keyword filters on. You can count, you can put kind of crude, crude keyword filters sort of on top, but you can't guarantee that it's not going to you talk abo…”
Evans: TikTok's algorithmic flow prevents creators from building a permanent owned audience.
“I don't get the impression that TikTok has that same mechanic of, like, I can build a permanent audience there, and that it's more that you are more dependent on the river, on being thrown up to the river, because, you know, TikTok, YouTube gives you all these…”