Attempts to use AI for "synthetic customer calls" failed because the models are overly agreeable, expressing a 10/10 purchase intent for any idea. This "sycophancy mode" makes them useless for genuine validation, proving there is no substitute for talking to real, nuanced humans.
The incubator focuses on starting one company every two years, running it to $5-10M revenue, then hiring a CEO to scale. This model allows the founding partners to specialize in the difficult 0-to-1 phase while retaining significant involvement and ownership.
Boulton & Watt built an internal AI agent that processes customer interview transcripts. It maps findings to core hypotheses, highlighting supporting and contradicting evidence. This keeps the team rigorous and fact-based, counteracting natural founder bias during the discovery process.
Creating an "AI initiative" can be a mistake, as it encourages tool usage for its own sake. A better approach is to set the expectation that team members will deliver the best possible outcome, knowing AI exists, shifting the focus from process to high-quality results.
Instead of building AI-native companies facing intense competition, a viable strategy is to build "AI-durable" businesses. These are in real-world sectors (e.g., funeral homes) where the core service isn't disrupted by AI, but operations can be significantly accelerated by it.
The impact of AI on engineering productivity is not uniform. For new, greenfield projects, seed-stage founders report up to 10x speed improvements. For established companies with mature codebases (e.g., Series D), gains are much more modest, around 10%, due to integration complexity.
Founders often feel existential dread in years 4-10 as a company shifts to pure execution. The Boulton & Watt incubator model sidesteps this by having partners transition out of the CEO role after the initial creative phase, allowing them to focus on what they enjoy most.
The founder journey requires different skills at different stages. Instead of being a generalist CEO for ten years, founders can specialize in the chaotic 0-to-1 phase. By repeatedly building companies to initial traction and then handing them off, they get more reps and build deep expertise.