Eliot Prince

One of my community members spent 5 to 6 hours editing every podcast episode. Now it's 20 to 30 minutes, and AI flags weak hooks by comparing each cut against her best past episodes. She spends the time she got back on ideas and angles. Not clicking buttons in Premiere Pro.

Most people abandon their first Claude skill fast. It's usually one of 3 mistakes: 1. Skipping the run-through. Do the job together once, then say "turn this into a skill". 2. One skill doing five jobs. I run 15 SEO skills, one job each. 3. A vague description. Claude reads it to decide when to use the skill, so spell out the trigger. Train it like an intern before you hand over the keys.

Most people using AI are doing it wrong. They're buying tools instead of building a system. The system is boring: Your business, written down Your repeat jobs, turned into skills Your apps, connected Your routine work, scheduled It still works when next month's tool wins.

If your washing came out dirty, you wouldn't bin the machine. You'd check the cycle, the temperature and the soap. Yet that's how most people judge AI. One go on the free plan, one generic answer, and the verdict is "rubbish tool". The first attempt at anything useful often takes longer than doing it yourself. That's exactly where most people give up. I tell students to treat AI like a whip-smart intern on their first day: sharp and keen, but knows nothing about your business yet. You don't sack the intern after one bad draft. You show them how you'd do it. What tool did you try but thought was rubbish after one go?

It's never been easier to build your own software. Nobody mentions the maintenance tax. Build it for you? Great. Build it for dozens of people? You're now the one fixing it. I learned that running a system used by 3,000+ people meant a lot of my day became bug fixes and feature requests. Every new feature also opened up a whole new load of issues. Just because you can build it doesn't mean you should build it. Or at least, be ready for the maintenance tax.

One of my community members used to spend hours on eBay listings. or in fact, not even bothering because it was such as chore. Now it takes under 5 minutes. He drops the photos into Claude and a skill does the rest: - Identifies the item - Checks the condition - Researches the price - Writes the listing - Publishes it No clever prompting involved. He wrote the checklist once and turned it into a skill. Boom, job done. Money made, life better.

I asked ChatGPT who it recommends for AI training, and it named four local training companies. None of them was me, lol 😅 Pretty embarrassing as a recovering SEO specialist. The ones it recommended had deep information about their services on their websites, across multiple pages. Every question a person might want to know is answered. My site missed dozens of these gaps. AI can't recommend you if it doesn't know what you do.

A prompt is something you do once. Anything you do weekly should be a skill.

Most businesses are still running a month behind, using last month's report, printed out and discussed in a meeting. They're running their business on trailing, outdated indicators. Connect your data to Claude and you get a live dashboard instead: what to do today, not what should have been done a month ago. One member of my community built his dream dashboard and called it "an absolute relief".

Nobody has 10 years' experience in AI. ChatGPT came out in 2022. The best head start anyone has on you is a couple of years, and most are still only using it to rewrite emails. You're not far behind.

Every Claude chat I opened was carrying a 65,000-token burn before I typed a word. Yours is too. Type /context into any chat and it opens up a secret portal. Mine: 20,000 on the system prompt 45,000 on the connectors Plus a whole load more Try it for yourself. Find all the token burn and switch off what you're not using.

Bored this weekend and want to make more money? Try this Claude + Free Money method… Everyone thinks the only way to make more money is to sell more. Not always. There's money sat on the table for businesses like yours already. Grants, wage subsidies, funded hires, tax reliefs and free credits. Plenty of owners are too busy to look, so they assume they don't qualify. You just can't dig through it all in the time you've got. Claude can though…. The game plan: 1. Make sure Claude knows your business Switch on research. Then give it something to work from: your website link will do. — 2. Run this prompt: "You are Martin Lewis, founder of MoneySavingExpert, applying your 'never assume you don't qualify' rule to business funding. Find me grants, wage subsidies, funded hires, tax reliefs and free credits my business is likely to get, so I stop leaving free money on the table while I'm too busy to look. Build my profile from everything you already know about me (memory, past chats, files, website) and only ask about gaps that change eligibility, especially location and legal structure. Search live, local first then national and sector, and verify every scheme on the funder's own page. Give me a ranked table of up to 10 open opportunities with amount, deadline, fit, effort, the catch and official link, then the 1 to 3 quick wins to action this week." — 3. Get back a ranked table of up to 10 open opportunities Amount, deadline, fit, effort, the catch and the official link for each. Then the 1 to 3 quick wins to action this week. When I ran it, it found a local innovation grant I could apply for. Ways to hire and train young people for almost nothing. And even sneaky tax breaks I never knew existed. I had no idea… It even went ahead and filled in my grant application for me. I'll probably come out at least £10k better off from this one 🤑 Before you chase the next sale, check what's already on the table. Ask Claude to find it.

Looking for the best AI-ish in person events, conferences, masterminds to attended over the next 12 months… What’s out there? Where am I heading?

If your work is good enough to grow someone else’s business, it’s good enough to grow your own

7 Books to Read While Claude Works (that Make You Impossible to Beat) 1. The Dip — Seth Godin 2. 10x Is Easier Than 2x — Benjamin Hardy 3. Reframe Your Brain — Scott Adams 4. Atomic Habits — James Clear 5. Oversubscribed — Daniel Priestley 7. $100M Offers — Alex Hormozi

three things that give me the digital ick: - faceless youtube channel (junk food of the internet) - side hustle (make it the main mission, not a side piece) - passive income (why you so scared of work?) also the word synergy is disgusting. ✌️

SLOP = Superb Lack Of Personality ✌️🤖

Three steps to learning AI: 1. Pick an AI tool (Claude) 2. Spend 2 hours leaning every feature 3. Pick a problem in your life and solve it with AI That will be the light switch moment when you understand what all the hype is about.