Summary

A conversation between Steven Bartlett and entrepreneur Daniel Priestley about where money is still to be made as AI collapses the cost of building software and making content. The useful tension is that Bartlett repeatedly presses the obvious objection — if everyone can do a thing cheaply, that thing stops being worth anything — and the answer they arrive at is that the product is no longer the moat. Posted on the channel's clips feed rather than the main channel.

• The worked example, which is the concrete part: Bartlett's company had been paying tens of thousands a year for an applicant tracking system, and rebuilt it in-house in about a week. Priestley estimates the same build would previously have cost around £500,000 and taken 18 months. Bartlett says the bespoke version is already better than what it replaced.
• Priestley's argument is that small SaaS has stopped being elite. Where a software company once needed 20-30 developers, millions in funding and around 10,000 customers to break even, he says there are now profitable software businesses serving 500 to 1,000 customers in a narrow niche.
• Bartlett states the counter-argument himself and does not resolve it comfortably: the same collapse in cost that lets him build an ATS in a week means 'there's going to be a thousand new ATS systems a day'. Value falls at the same rate that difficulty does.
• Their answer is that the software is no longer the product — the ecosystem is. A tool bundled with training, a community, a bootcamp, a retreat or a dinner series is defensible in a way the tool alone is not, because the tool can be replicated in a day and the community cannot.
• The blue ocean timeline has compressed, in Priestley's account, from roughly ten years to months: 'as soon as we've identified that you've got a blue ocean, we can replicate that real fast'. His example is social media agencies, uncontested in 2014 and crowded by 2024.
• The 'irreplaceably human' test they both apply is what an AI cannot do rather than what it does badly: turn up to an event, stand on a stage, host a dinner, or speak from lived experience. Priestley's formulation is 'relatable beats impressive'; his example is a financial planner whose TED talk about meeting a hundred lottery winners drew half a million views because nobody else could tell it.
• On money: they argue the earner is a product and service ecosystem rather than one product, and Priestley estimates Bartlett has around 27 separate income streams — speaking, podcasting, AdSense, sponsors, shareholdings, coaching, book royalties, software.
• Two figures given without sourcing and flagged here as unverified: that legacy legal tech and data firms have lost roughly 20% of their value in 2026 with $280bn wiped off publicly traded companies in one week, and that Spotify's best developers have not written a line of code since December. Bartlett introduces the second as something he read.
• The legal anecdote is first-hand: a case quoted at around £50,000 to begin with a law firm, resolved instead using Claude on a $20/month subscription, which produced decision-tree options, the required documents and a negotiation script. Their conclusion is not that lawyers disappear but that the billable-hour model for regurgitating contracts does.
• Priestley's closing argument is a reversal: university-loan policy pushed people away from trades into degrees, and the resulting shortage means the blue ocean is now plumbing and electrical work, with tradespeople potentially out-earning lawyers.

Why it matters

This is the batch's only entry about what an individual should actually do, and its central claim is one the Handbook can test over time: that as production cost falls to nothing, value moves to whatever cannot be produced — presence, relationship, and lived experience. The ATS example is the strongest evidence in it, because it is first-hand, recent, and specific about both cost and duration. Treat the £500,000 and 18-month comparison as an estimate made in conversation rather than a costed figure. The two unsourced statistics should not be repeated without checking. Note also that the advice and the interest are not independent: both speakers sell products in the ecosystem model they are recommending.

HKWLH1Zh79A-transcript.txt