AI Automation in Boring Industries

What It Is

Look for a recurring, painful workflow inside an ordinary 'boring' business still run by phone calls, email, and spreadsheets, since cited data shows most small businesses have started experimenting with AI but very few have built it into core operations (a Goldman Sachs survey found 76% using AI but only 14% had integrated it into operations). Rank candidate opportunities across seven areas by ease of starting, value of the problem, and required expertise, from easiest (local marketing/lead follow-up, home services, property management, small business bookkeeping) to more specialized areas requiring real domain expertise. Start at the bottom of the list if new to this; lean into existing industry experience (accounting, logistics, legal, healthcare) if you have it, since that expertise becomes your advantage further up the specialization ladder. For your first client, start with someone you already have access to, ask them to walk you through their messiest recurring process, and build a pilot around the one complaint that comes up repeatedly. Price based on value delivered, not hours worked. Run every opportunity through a 5-step, 90-day test: pick an industry you understand, map one full customer journey to find manual wait/copy/follow-up points, find what the business already spends on that problem, package one clear before-and-after result, then give it 90 days — needing three yeses: did anyone pay, can you repeat the result, can you deliver it faster for a second client.

The Pitch

The presenter draws on more than 50 real interviews from her own podcast with named AI founders and CEOs (including Higgsfield's Alex Mashrabov, Replit's CEO, Coursera founder Andrew Ng, Decagon's Jesse Zhang), citing a concrete example: an accounting firm using AI to identify a client qualified for an R&D tax credit, then using AI-assisted paperwork to help them claim roughly $50,000 in tax credits, recurring every year.

Who It’s Best For

Someone willing to do real customer discovery rather than guessing at a problem to solve, comfortable running a genuinely time-boxed 90-day test rather than committing indefinitely upfront.

The Truth (as we see it)

The cited statistics (Goldman Sachs survey, OECD adoption data, an Andreessen Horowitz analysis of 600+ industry categories) are real, independently sourced data points, and the interview quotes are from real, named, identifiable industry figures. The R&D tax credit case study is specific and concrete though relayed secondhand. The video includes a paid sponsorship segment for an email marketing analytics tool, separate from the core research-driven content.

Our assessment, based on the video and what’s typical across similar videos. We don’t independently verify income claims or prices.