Feb 23, 2026 · 31m · another-podcast
AI and SaaS
0:00 / 0:00
gold bands on the timeline = statements, start to end. Hover to read, click to jump. CC turns on captions
Tony Cameron and Benedict Evans analyze how artificial intelligence is reshaping enterprise software, arguing that while AI drastically lowers coding costs and automates routine workflows, long-term value remains anchored in deep product design and nuanced human judgment.
How this conversation actually went
Every chapter scored 0–10 on four independent dynamics. Hover any point for the reasoning behind the score. The hosts hold 80.9% of the talking time here. How this is scored →
The hosts as informed peer 7.1 Guest teaching 1.6 Guest disagreement 1.4 The hosts pushing back 2.0
speaking balance: gold is the hosts, purple is the guest (3 minute bins)
Sharpest disagreement ▶ 19:03 Cameron challenges Evans's view on worker optimization
Cameron directly challenges the premise that workers should not optimize their workflows, arguing that people should naturally want to do more with less.
Hardest push from the hosts ▶ 13:04 Evans counters Cameron's growth framingEvans immediately rejects Cameron's suggestion that automation and scale growth are simultaneous, clearly redirecting to his stepwise transition framework.
Biggest teaching moment ▶ 28:24 Cameron breaks down tacit F1 paddock knowledgeCameron educates Evans on the distinct difference between AI-prompted sports writing and firsthand paddock knowledge working directly with race engineers.
The host holds their own ▶ 16:07 Evans articulates the abstraction shift in enterprise queriesEvans demonstrates deep conceptual mastery by contrasting deterministic database queries with probabilistic predictive insights in enterprise systems.
the scores for every segment, with the reasoning behind each
| Chapter | Topic | The hosts as informed peer | Guest teaching | Guest disagreement | The hosts pushing back | Why |
|---|---|---|---|---|---|---|
| Waves of AI and Market Psychology | 8 | 1 | 1 | 2 | Benedict Evans commands the discussion with historical parallels to the dot-com bubble, cloud economics, capex dynamics, and software abstractions. Tony Cameron acts primarily as an agreeable conversational prompt, validating Evans's analysis. | |
| The True Challenge of Software Creation | 7 | 1 | 1 | 2 | Evans dismantles the assumption that coding is the primary barrier in software development, elaborating on Paul Graham's tar pit concept and the historical transition to cloud architectures. Cameron chimes in supportively with brief observations on on-prem transitions. | |
| Predictive Intelligence and Higher Abstraction Layers | 8 | 2 | 2 | 3 | Evans outlines a three-step platform adoption framework and uses Walmart store management to illustrate higher abstraction layers. Cameron attempts a brief interjection on automation vs growth at scale, but Evans quickly clarifies his specific progression model. | |
| Workplace Optimization and User-Driven Tooling Fallacies | 6 | 2 | 3 | 4 | A friendly philosophical debate emerges over whether normal employees should optimize their own workflows. Cameron advocates for innate curiosity and doing more with less, while Evans pushes back by noting that product management is a specialized discipline. | |
| Workflow Complexity and the Frame.io Case Study | 7 | 1 | 1 | 1 | Evans draws on his venture capital experience with Frame.io to detail how complex human workflows resist simplistic ad-hoc tooling. Cameron listens attentively as Evans reinforces the distinction between coding and product workflow design. | |
| Creative AI, Authenticity, and Canned Music | 7 | 2 | 1 | 1 | Cameron introduces the public backlash against creative AI, and Evans extends the thesis by referencing historical resistance to recorded music and photography. Both speakers build collaboratively on the concepts of intent and authenticity. | |
| Domain Judgment vs. Boilerplate Generation | 7 | 2 | 1 | 1 | Cameron shares domain-specific insights on Formula One journalism and paddock expertise, demonstrating where generic AI generation fails. Evans ties this back to management consulting and commodity software wrapping. |
Statements from this episode (16)
Evans: Replacing enterprise software like SAP with Claude is foolish
“Everyone who works in enterprise software and knows anything about it, kind of put their head in their hands and said, oh my God, do I actually have to explain why this is dumb? Is, you know, you'll use Claude instead of SAP. Fine. If you need me to explain th…”
Evans: Big Four tech capex will jump from $400B to $650B
“The big four companies, platform companies, spent four hundred billion dollars on capex last year. This year, they've said they'll spend 650, more or less. Microsoft hasn't been given full year guidance. But Meta and Google and Amazon have more or less doubled…”
Evans: Private equity holds $500B in debt from software buyouts
“There's a big piece in the FT this morning about how much money the private equity industry put into buying software companies in the last 10 or 15 years, and a whole amount of debt, like half a billion, half a trillion dollars of debt around that.”
Evans: The machine learning wave ten years ago was not transformative
“I mean, the last wave of AI was machine learning, 1015 years ago now, 10 years ago, really. That was not transformative across everything. It was a new bunch of stuff that everybody could build, and there were a bunch of new companies with it, but it wasn't th…”
Evans: Dismissing AI as useless is like calling the internet useless in 1999
“I still see people who say, no, it's just like a stochastic parrot. It doesn't do causation. It doesn't work very well. It's got hallucinations. It's all useless. These people are morons. Like, they are just idiots. But they're idiots in the sense of somebody …”
Evans: AI makes producing software orders of magnitude cheaper and faster
“The first of them is, well, clearly this is an order of magnitude, maybe several orders of magnitude, cheaper to make any given piece of software. A piece of software that you might have thought of 10 years ago that has nothing to do with AI, you can use AI to…”
Evans: The delusion that AI will turn non-coders into programmers is wrong
“There's a recurring delusion throughout the history of software that firstly, people always think there's going to be a general purpose abstraction layer that will just do everything. And secondly, people always think that everyone who doesn't write code will …”
Evans: The hard part of making software is almost never writing code
“The underlying point, though, is I think the hard part of making software is almost never writing the code.”
Evans: Cloud accounts for only a third of enterprise workflows
“It's still only sort of a third of enterprise workflows. So it takes a long time to re-platform.”
Evans: Oracle raised $50 billion in bonds this year for cloud
“Borrowed a lot, an enormous amount of money, fifty billion dollars of bonds this year to capital raising this year.”
Evans: Mandating AI Usage Is a Fallacy Like 1998 Internet Mandates
“There's a big, fuzzy question of how much do you, as a normal person working for a company, have to be thinking about what the software could be doing for you, which I think gets you into this fallacy of, like, we're going to make everybody use AI, which to me…”
Evans: AI works best where critics are angriest about it
“The fields where AI is actually works best right now are the fields where people are really angry about it and sure it's useless. ... Whereas the fields where it doesn't, really doesn't quite work, or needs an awful lot of other stuff around it, The fields whe…”
Evans: Creators most upset about AI art lack originality
“The really unkind thing one could say is that the people who are most upset about AI art are the people who are making stuff that has no particular authenticity originality. I mean, if you, if you've made, if you wrote 10 vampire romance books last year and yo…”
Evans: McKinsey clients pay for strategic problem solving, not slides
“You have no idea what McKinsey does or what the clients are paying for. They're not paying for the slides.”
Evans: I will not publish writing that ChatGPT could generate
“One of my, the ways I would think about what I'm writing is I've written this. Is this what anybody else would say? Is this just saying what everybody knows or everyone thinks? Well, then I won't publish it. And now I just say, well, is this what ChatGPT would…”
Benedict Evans: SaaS that merely wraps SQL with custom branding is doomed
“If you've got a piece of software, all you're really doing is wrapping up SQL and putting a car logo on it instead of a bike logo on it so that it's car software instead of bike software, then yeah, you're screwed. That's what happened with cloud too.”