--- title: "Dangerous language in startups" author: "Matt Brown" canonical_url: "https://notes.mtb.xyz/p/dangerous-language" source: "notes.mtb.xyz (Substack)" year: "2024" primary_topic: "frameworks" tags: ["frameworks"] fetched_at: "2026-05-14" note: "Markdown twin of the canonical Substack post. The canonical URL above is authoritative; this file may lag if the original is edited." --- ### How we describe something affects how we think about and interact with it. It's especially potent and dangerous in startup descriptions, and worth remembering for the companies building "with AI". I think about the book [Metaphors We Live By](https://press.uchicago.edu/ucp/books/book/chicago/M/bo3637992.html) at least once a week. The TLDR is that metaphors aren’t just abstract or poetic expressions. Instead, they consciously and unconsciously shape how we think about and interact with the world. Some fun examples: “argument is war” (e.g., we attack a position, defend our argument, win or lose a debate), “time is money” (e.g., we invest, spend, save, or waste time), and “ideas are food” (e.g., we digest ideas, find others hard to swallow). ![Book cover for "Metaphors We Live By" by George Lakoff and Mark Johnson, with a textured purple-and-grey checkerboard pattern and a central purple circle above the title set on a dark background, with "With a new afterword" below the authors' names.](https://substackcdn.com/image/fetch/$s_!phXf!,w_1456,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F5c5a7665-48c7-4804-a1a2-71a5e55d9b87_860x1363.jpeg) Metaphors We Live By The larger lesson is that **the language we use to describe something affects how we think about and interact with it**. There are many interesting applications of this, and one close to my day job as a VC is the language startups use to describe themselves. It’s not easy to compress a company’s vision, market, and strategy into something that’s both concise and compelling. That’s why founders fall into well-worn patterns. “X for Y” is one such pattern. It’ll be nostalgic for anyone who was in startups in the early 2010s. Every time you refreshed TechCrunch there was another funding announcement about the “Uber for power tools” or “Airbnb for RVs” or “Linkedin for blue collar workers”. ![Collage of old TechCrunch headlines including "Doximity, A LinkedIn For Medical Professionals," "Talent Maven Launches As A LinkedIn For Actors, Musicians, And Models," "Iraq Veterans Launch RallyPoint: A LinkedIn For The Military," "Uber-For-Beauty Startup StyleBee," "Uber-For-Laundry Startup Washio," "BlackJet, The Uber Of Private Jets," "HomeDine, The Airbnb For Home-Cooked Meals," and "'Airbnb For Bikes' Startup Spinlister."](https://substackcdn.com/image/fetch/$s_!17L2!,w_1456,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Ff8c65999-d56d-423e-9d37-5e7b7c61926c_1600x948.png) A sample of TechCrunch headlines from the “X for Y” era It’s a powerful shortcut. It purposefully associates a new company with the success and size of an existing one. This helps to explain and impress. But it’s also dangerous. The “X” in the equation does a lot of work. It suggests that the market size, dynamics, and strategies of X apply to Y. It focuses attention on an already-built thing (X), rather than thinking from first principles of the problems and opportunities of the new market (Y). Even though it seems innocent, the metaphor shapes how founders, employees, and investors make decisions in and about the company. A similar pattern is increasingly popular today, and it’s no less dangerous if used incorrectly. I call it “with X”. And nowadays, X is AI: “marketing automation with AI” or “fraud prevention with AI” or “medical scans with AI”, etc. ![Collage of TechCrunch headlines from the AI era, including "Aisera lands $90M to automate customer service requests with AI," "Source.ag raises $23M to raise the bar on raising crops with AI," "India's Kombai raises $4.5M to simplify UI coding with AI," "Oscilar emerges from stealth to fight transactions fraud with AI," "Seadronix aims to reduce marine accidents at port and sea with AI," and "Inscribe bags $25M to fight financial fraud with AI."](https://substackcdn.com/image/fetch/$s_!P_E_!,w_1456,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F1c9066f1-cca9-4109-994b-a816d3be0887_2082x960.png) A sample of TechCrunch headlines from the “with X” era I’m far from an AI skeptic. I think it’s the biggest tech advancement of this decade, if not century. I’m a daily user of multiple AI products and actively invest in companies using AI. The problem is that this shorthand can warp how founders and investors think about these companies. Like the “X for Y” pattern, the “with AI” pattern deliberately associates a company with something exciting and successful, a massive technology shift. It does the job of getting attention and building excitement. But that tradeoff for conciseness can be dangerous. **Just like “X for Y” patterns harm founders by focusing them on X rather than Y, the “with AI” pattern can focus founders on the AI rather than the problem being solved.** Does this mean that you shouldn’t use AI as a startup? Or that you shouldn’t brand yourself as using it if you do? Absolutely not. It’s a powerful technology and marketing tool. **But know that your framing can shape your thoughts and actions.** There’s the popular saying: “People don’t buy a quarter inch drill. They buy a quarter inch hole.” It’s a good reminder that customers ultimately aren’t buying the thing to solve the problem, they’re buying a solved problem. Those two things are often but not always the same. **This is worth reiterating, since the tech industry fetishizes the new thing. What’s tricky is that this fetishization of the new thing often generates early traction.** Your first 10 customers may use your product, not despite, but *because* of the “with X”. Most of the early desire to use something may be because it uses AI – not necessarily because it solves the core problem better with it. And that can be enough to get you going, so it shouldn’t be ignored. But the effect lessens over time. Your first 10 customers really care about AI. Your 100th sort of cares about it, but really cares about how much better you solve their problem. And your 1,000th customer is probably skeptical of AI and just wants the problem solved. It reminds me of the classic “Crossing the Chasm” model, as a company transitions from the early adopters to the early majority: ![Hand-drawn diagram combining the Crossing the Chasm bell curve (Innovators, Early Adopters, "The Chasm," Early Majority, Late Majority, Laggards) with a stacked area on top labeled "Relative importance of HOW a problem is solved" (blue, large on the left) and "WHAT problem is solved" (green, dominant on the right), illustrating that earlier customers care more about how a problem is solved and later customers care primarily about what is solved.](https://substackcdn.com/image/fetch/$s_!g4qG!,w_1456,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2Fd8348d4e-375c-49b9-901f-6c8a7cc78c79_1600x1092.png) The challenge for founders is to take advantage of the “with X” framing in the early days, without letting it warp their thinking and strategy too much. This means staying focused on *what* problem is being solved and not *how* it’s solved. It means remembering that AI, or any other technology or business model, is a means to an end. It means remembering that the overwhelming majority of your future customers won’t know or care about *how* you solve a problem, but *what* problem you solve and how much better/cheaper/faster you do it for them. That can be difficult because of the power of language. How you describe the company informs how you think about the company. How you think about your company informs, consciously and unconsciously, in big and small ways, the decisions you make. It reminds me of the [famous scene](https://www.youtube.com/watch?v=0gU35Tgtlmg) from *Indiana Jones and Raiders of the Lost Ark*, where Indiana has to switch a boobytrapped idol with a bag of sand. Like in the scene, founders must use the power of the current thing (“with X!”), but swap it in their own minds and eventually in the market with the actual problem that’s being solved. They must time this correctly, so they don’t cling on to *how* a problem is solved at the expense of actually solving it for the greatest number of people in the long run. ![Still from Raiders of the Lost Ark of Indiana Jones reaching across a pedestal to swap the golden idol with a bag of sand, with a hand-drawn blue sticky note over the sandbag labeled "…with AI!" and a green sticky note over the gold idol labeled "Real problems, big markets."](https://substackcdn.com/image/fetch/$s_!hSig!,w_1456,c_limit,f_webp,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F690a8f98-6db4-4390-bb04-28b863080677_1600x852.png) Smart founders recognize that the goal isn’t simply to build a company using AI – or any other specific technology or business model. The real goal is to build a large and successful company, and AI (or any other technology or business model) is just a means to an end. “With X” is a gambit to build momentum until you’re better known for solving a problem really well than for how you solve it.