Aug 16, 2026 · 25m · another-podcast
AI deployment
0:00 / 0:00
gold bands on the timeline = statements, start to end. Hover to read, click to jump. CC turns on captions
Benedict Evans and Toni Cowan-Brown examine the real-world operational challenges of enterprise AI adoption, contrasting Silicon Valley's disruptive narratives with the slower realities of corporate change management, procurement, and historical technology cycles.
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 99.9% of the talking time here. How this is scored →
The hosts as informed peer 7.7 Guest teaching 1.7 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 ▶ 15:05 Pushing Back on In-House Software Builds
Toni emphatically rejects the idea of non-tech organizations building custom software in-house by citing presidential campaign tech failures.
Hardest push from the hosts ▶ 22:47 Challenging the Autopilot FramingBenedict interrupts Toni's setup on 'copilot versus autopilot' to directly question what autopilot actually means in practice.
Biggest teaching moment ▶ 11:52 The Limits of Bottom-Up SaaS AdoptionToni clearly outlines how bottom-up SaaS breaks down when cross-departmental, end-to-end organizational change is required.
The host holds their own ▶ 23:07 Copilot as a Fragmented Corporate BrandBenedict demonstrates deep industry insight by breaking down Microsoft's internal PM incentives and comparing Copilot's disconnected product suite to IBM Watson.
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 |
|---|---|---|---|---|---|---|
| Framework for Technology Adoption and Early AI Pilots | 8 | 1 | 1 | 1 | Benedict establishes a clear three-part analytical framework for enterprise tech adoption, drawing detailed historical parallels between industrial companies like Caterpillar and newspapers during the rise of the web. | |
| The Illusion of Leverage and the AI-Enabled Label | 8 | 1 | 1 | 1 | Benedict deconstructs the term 'AI-enabled company' by drawing analogies to legacy banks versus neobanks and airline booking systems, showing deep historical domain expertise. | |
| Change Management Realities Versus Silicon Valley Assumptions | 8 | 2 | 2 | 3 | Toni questions whether labeling things 'AI-enabled' stems from conceptual confusion. Benedict immediately counters with change management literature from the 1980s and mocks extreme AI doomer assumptions regarding enterprise replacement. | |
| Enterprise Inertia and the Limits of Bottom-Up Adoption | 7 | 3 | 1 | 2 | Toni articulates why bottom-up SaaS adoption models fail when systemic change management is required, prompting Benedict to validate and expand on enterprise procurement timelines and integration complexity. | |
| Build Versus Buy Decisions and Historical Parallels | 8 | 2 | 2 | 2 | Benedict references early 1980s computing history, including Boeing's proprietary 3D spreadsheet program, to illustrate historical build-versus-buy cycles and commodity software dynamics. | |
| Procurement Mechanics and the Fragmented Reality of Copilots | 8 | 2 | 2 | 4 | When Toni introduces the copilot versus autopilot distinction, Benedict immediately challenges the term autopilot and delivers a sharp critique of Microsoft's fragmented branding of Copilot. | |
| Unbundling Enterprise Workflows and the Enduring Role of Consultants | 7 | 1 | 1 | 1 | Benedict synthesizes how enterprise startups and management consultants serve complementary roles in identifying unbundled operational workflows. |
Statements from this episode (10)
Evans: Giving enterprise employees Copilot or ChatGPT has not worked well
“Step one was give everybody co-pilot, and that didn't work too well, because it's kind of, or Claude or ChatGPT, because it's kind of like being in 1997 and giving everybody a web browser. Okay, this is the future, but it's not, that's not, what was the produc…”
Evans: Calling companies 'AI-enabled' resembles 'internet-enabled' branding in 1998
“Or the end of the day is saying that we're an AI-enabled XYZ, kind of like saying we're, you know, we're an internet-enabled law firm in, like, 1998.”
Evans: Enterprise tech adoption poses management challenges rather than engineering hurdles
“Clearly the deployment of mainframes and data processing and spreadsheets and ERPs and so on Did kind of completely change the economics of your business, and that came with a whole bunch of challenges, most of which were really internal management challenges,…”
Evans: Ubiquitous enterprise tech efficiency gains get competed away in price
“We'll all get some cost savings out of it and some efficiency gains that will get competed away in price because we're all doing it. And at the end of the day, as long as you don't screw it up, this doesn't change the competitive dynamics. Very much.”
Evans: Incumbent industrial firms will not be displaced by AI-accounting startups
“Saint-Gobain is a big French gloss company. They're going to be destroyed by a new glass company that's using Claw to do its accounting. You know, it's a reductive absurd, but of course we'll do this absurd.”
Cowan-Brown: Bottom-up SaaS fails when value requires change management
“It lands with a team, you expand, and that works when the tool is replacing or when the tool is making an individual better at the job that they already have. It fails and it gets complicated when the value requires real actual change management.”
Evans: Enterprise sales and integration cycles take at least 18 months
“Jesus Christ, it takes three to six months to get somebody at an enterprise to reply to your inquiry about selling them some software. You know, the enterprise sales cycle is what, 18 months? At best? You know, you want to add a new thing and plug it into SAP …”
Evans: Corporate AI questions either lack answers or follow historical precedents
“All the questions I get are going to fit into two categories. The question, the answer is either no one knows, or how did it work last time?”
Evans: Microsoft Copilot is an umbrella brand of disconnected features like IBM Watson
“Copilot is rather like Watson. It's a brand name that represents an, that represents every single PM at Microsoft had to add an AI feature as part of their quarterly goals. And so every, every third button at Microsoft And every Microsoft product is purple and…”
Evans: The idea that AI ends management consulting is unintentionally hilarious
“It's unintentionally hilarious that people think this is the end of consultants. It's kind of a point I've come back to several times. You know, nobody is going to work this out by themselves, and most people are going to pay somebody to do it for them.”