Where AI actually lands, industry by industry
AI works when it fits the sector constraint. This piece maps twelve sectors, the useful applications, and where we stop.
ReadInsights
Category: Product
Deciding what to build before deciding how. Discovery, scope, pricing models and the product calls that turn out to matter most later on.
Open Lance, what taking no commission actually did
What happened when we removed commission from Open Lance, including the assumptions that held and the ones users proved incomplete.
ReadThe agent marketplace thesis, in full
BotUp argues AI work should be bought as defined jobs with specs, prices, evaluation and remedies, rather than as access to a model.
ReadWhat AI does to freelance work, and what it does not
AI has cut some freelance categories, grown others, and changed many more. The split depends on whether buyers can judge the work.
ReadReputation you can take with you
Marketplace reviews keep workers tied to a platform. Portable records need verified facts, signatures, attribution and revocation.
ReadDisputes, the feature no marketplace wants to design
Disputes need structured agreements, timers, defaults, partial outcomes and human accountability when money is on the line.
ReadMoney that has already cleared, and the five days you wait for it
Freelancers wait for cleared money because holds mix real risk, weak escrow design, and float that benefits platforms.
ReadZero commission is a business model, not a discount
Zero commission moves the price into verification, placement, tooling and organisation software, with harder engineering behind it.
ReadAI breaks the leverage model professional services is built on
AI is compressing junior work in professional services, so the old pyramid, training model and hourly pricing no longer hold.
ReadEvery marketplace is a trust machine wearing a fee
Marketplace fees pay for trust, but percentage commissions often outgrow the cost of escrow, discovery, disputes and reputation.
Read
