Ally Mac

Helping ordinary people use AI to create profitable businesses by unlocking what they already know.

[Ally Mac @allymac81]

If you manage a team, review season means hours staring at a blank document trying to say something coherent about people you know really well.

Pull your 1:1 notes from the past 6 months, drop them into Claude or ChatGPT, and ask it to draft the review. Spend 20 minutes editing.

Managers in HR subreddits have been doing this for over a year. The consistent report is they used to dread reviews. Now it's just editing a draft.

[Ally Mac @allymac81]

If you manage a team, review season means hours staring at a blank document trying to say something coherent about people you know really well.

Pull your 1:1 notes from the past 6 months, drop them into Claude or ChatGPT, and ask it to draft the review. Spend 20 minutes editing.

Managers in HR subreddits have been doing this for over a year. The consistent report is they used to dread reviews. Now it's just editing a draft.

[Ally Mac @allymac81]

Heather Gauck is a special ed teacher in Grand Rapids. She uses AI to adapt reading passages to different levels. What used to be hours of manual rewriting is now seconds.

She drafts parts of her students' IEPs with it. She says she reclaimed about one full planning day per week. Same kids. Same classroom. Just less time on the stuff that isn't teaching.

A 2025 Gallup study found teachers using AI weekly save about 6 hours a week. That's 6 full weeks back every school year.

No course. No system. She just started asking.

[Ally Mac @allymac81]

If you manage a team, review season means hours staring at a blank document trying to say something coherent about people you know really well.

Pull your 1:1 notes from the past 6 months, drop them into Claude or ChatGPT, and ask it to draft the review. Spend 20 minutes editing.

Managers in HR subreddits have been doing this for over a year. The consistent report is they used to dread reviews. Now it's just editing a draft.

[Ally Mac @allymac81]

@gregisenberg agents are more predictable customers than humans. this is a feature not a bug for builders

[Ally Mac @allymac81]

@BoringBiz_ quitting $400k for pre-revenue anything requires a very specific kind of faith

[Ally Mac @allymac81]

@marclou hardest part is you can't know in year 3 if you're 7 years from winning or just 7 years from failing

[Ally Mac @allymac81]

@thejustinwelsh the 89% margin is the story. revenue without margin is just expensive stress

[Ally Mac @allymac81]

@regent0x_ isolation is the unlock. everyone builds one big smart agent and wonders why it breaks

[Ally Mac @allymac81]

Heather Gauck is a special ed teacher in Grand Rapids. She uses AI to adapt reading passages to different levels. What used to be hours of manual rewriting is now seconds.

She drafts parts of her students' IEPs with it. She says she reclaimed about one full planning day per week. Same kids. Same classroom. Just less time on the stuff that isn't teaching.

A 2025 Gallup study found teachers using AI weekly save about 6 hours a week. That's 6 full weeks back every school year.

No course. No system. She just started asking.

[Ally Mac @allymac81]

nurses are still spending 1-2 hours after every shift finishing their charting.

AI medical scribes that listen to patient encounters and auto-generate clinical notes have been available for a couple of years. some hospitals have rolled them out. many have not gotten there yet.

a nurse who finds and starts using a tool like Nabla or https://t.co/1So8xapoTQ gets her evenings back years before her hospital IT team builds a rollout plan.

between when something works and when an institution officially deploys it, there is a window where individual initiative pays off.

[Ally Mac @allymac81]

the typical pre-call prep is one of two things. scrambling through notes right before joining, or going in cold and hoping the prospect fills the first five minutes.

there is a better version that takes four minutes.

dump everything you know about the prospect into Claude or ChatGPT. company, role, their situation, the goal of the call. ask it "what are three questions I should ask?"

you show up with real questions instead of "so tell me about your business." it changes how the whole call opens.

[Ally Mac @allymac81]

@askOkara ai gave everyone the tools. taste and judgment are the actual diff now

[Ally Mac @allymac81]

@gregisenberg weird niche one-person apps are the move and the guitar fruit ninja just proved it

[Ally Mac @allymac81]

@AnthropicAI if their internal team is at 8x, the gap between builders and people still watching is about to be very real

[Ally Mac @allymac81]

@jonoringer the AI stack he used is the actual story. 2 people doing what took 2,400 employees a few years ago

[Ally Mac @allymac81]

someone posted on Hacker News this month that they are running their one-person tech agency on Gemini free tier.

content creation, sales lead generation, security scanning, operations, all handled by AI agents they built themselves. live dashboard. agents running right now.

what stands out is not the tech stack. it is that one person, using free tools, is doing the operational work of a small team.

the question worth asking is whether the person competing with you is already doing this.

[Ally Mac @allymac81]

a teacher wrote for EdSurge this year about what AI did to her Sundays.

she used to spend four-plus hours every week before the school week even started. lesson plans, IEP notes, differentiated versions of readings for different ability levels, parent emails.

now she pastes a reading into ChatGPT and types "make this a kindergarten-level version." thirty seconds. she does the same for the other prep too.

she described it as getting "an entire planning day back."

Gallup measured this across thousands of teachers. the ones using AI weekly save 5.9 hours a week on average.

she did not learn a workflow. nobody built her a system. she just started asking.

[Ally Mac @allymac81]

most people leave a doctor's appointment with a printout full of jargon and a follow-up scheduled three weeks out. nobody explains what any of it means in the meantime, so people sit with it.

a lot of people have started pasting their lab results into ChatGPT and asking it to explain in plain English. you understand what is actually happening with your body in two minutes.

this is one of the most practical things AI does for ordinary people, and somehow it still feels like a secret.

[Ally Mac @allymac81]

@sama a clear playing field matters way more to small builders than people realize