Noah

Figuring out how to make real money with AI, in public. Simple side hustles, no fake screenshots. 🇬🇧

[Noah @NoahUK_]

I'm Noah. I'm using AI to build a side income from zero, no audience and no budget.

I'll share what works, what flops, and the real numbers as they come in.

If you want to make your first $1,000 online with AI, follow along.

[Noah @NoahUK_]

@TheSaaSCraftman recency is the whole trick. someone who felt the pain last week still remembers what it cost them

[Noah @NoahUK_]

@finnbuilt cold dms, but only to people already complaining about the problem in public. no guessing, and the first message basically writes itself

[Noah @NoahUK_]

@builtwithrubel D, and it's the one fully in your control. talk to 10 people before you build and most of A disappears with it

[Noah @NoahUK_]

@MartinsOsimen_ the experimenting part is the real filter. most people quit after the first thing that doesn't work, which is usually attempt one of about ten

[Noah @NoahUK_]

@konig0000 D, but sold as one narrow fix. "every enquiry answered in 5 minutes" for 3 local businesses at $350 a month is the whole $1k, no app needed

[Noah @NoahUK_]

Your first AI client doesn't need your website, logo or business name. They need to see the task done once, on their own stuff. Show it, don't pitch it.

[Noah @NoahUK_]

@sherifgjini useful for nudging an agent or fixing a typo while you're out. building anything real on a 6 inch screen is mostly a flex

[Noah @NoahUK_]

@DanteMMartin that's the real job in most service work. clients rarely ask the right question, the value is spotting what they forgot to mention

[Noah @NoahUK_]

@alexeheath @tobi if agents end up doing the buying, product pages stop being written for people. clear specs and honest reviews start beating clever copy

[Noah @NoahUK_]

@konig0000 selling custom builds is only overrated when every one is custom. build it once for one niche, sell it 20 times, and it's a different business

[Noah @NoahUK_]

Honest question for anyone selling AI services: what was harder, building the thing or getting the first person to pay for it?

[Noah @NoahUK_]

@xavier_coiffard yep. fix the list and the copy suddenly gets a lot easier to write

[Noah @NoahUK_]

@Fimiye fair enough lol. even 3 hours a week on one tiny thing still counts though. waiting for the perfect time is how it never starts

[Noah @NoahUK_]

@indie_alx not wrong. the $29 tools that survive will be the ones selling trust and support, not features. nobody wants to be on call for their own invoicing app

[Noah @NoahUK_]

@finnbuilt cold email for the first 10, because you hear exactly why they said yes. then wherever those 10 came from tells you where the next 90 are

[Noah @NoahUK_]

@mahaaaay a no with a compliment is still a reply. most cold emails never get that far

[Noah @NoahUK_]

@heyblake number 3 is the one people resist most. paying an agency feels like progress, posting yourself feels like exposure. year one usually needs the exposure

[Noah @NoahUK_]

Before you build an AI tool, do the job by hand for 3 customers. If nobody pays for the manual version, automating it won't change that.

[Noah @NoahUK_]

@Damir_Akaza auto-approve is fine on your own code. the moment the repo isn't yours, every readme and config file is basically a stranger typing into your terminal