5 tasks your AI assistant can handle while you sleep

Most people use AI as a fancy search engine — ask a question, get an answer. That's leaving most of the value on the table. A properly set-up AI assistant doesn't wait for you. It works overnight, delivers results in the morning, and never forgets.


Here are five things a well-configured AI assistant can genuinely handle while you're asleep — not theoretical capabilities, but real tasks that save real time every single day.

1
Sorting and flagging your inbox

An AI assistant can read your emails as they arrive overnight, sort them by priority, and give you a clean summary in the morning. You wake up knowing what actually needs your attention — not staring at 47 unread messages trying to figure out which three matter.

It can flag anything from a specific person, anything that mentions keywords like "urgent" or "invoice," and anything that needs a reply by a certain time. Everything else gets labeled and filed.

What this looks like in practice "Here's your inbox summary: 2 emails need a response today (a client follow-up from Marcus, and an invoice from your accountant). 14 newsletters have been archived. 3 receipts have been labeled. Nothing else is urgent."
2
Writing your daily briefing

Every morning, you could get a short brief covering exactly what you've asked for — today's calendar, any deadlines approaching, news about a topic you're tracking, or a summary of what you worked on yesterday.

This isn't a dashboard you have to go find. It's a message waiting for you when you wake up, written in plain language, covering only what's relevant to you. Takes zero effort on your part after the initial setup.

What this looks like in practice "Good morning. Today: call with Sarah at 2pm, proposal due Thursday. Three competitors launched updates yesterday — here's what changed. Your revenue this week is 12% ahead of last week. One thing worth knowing: your biggest client opened your last email but hasn't replied in 6 days."
3
Monitoring things you care about

Your AI assistant can keep watch on things you'd normally have to check manually. A competitor's website. Your own site uptime. A subreddit where your customers hang out. A keyword on social media. A price for a product you're thinking about buying.

Instead of you doing the checking, it does the checking — and only surfaces something when there's actually something worth knowing.

What this looks like in practice "Overnight alert: your main competitor changed their pricing page. They added a 'Starter' tier at $19/mo that didn't exist yesterday. Also: your site was unreachable for 4 minutes at 3:22 AM — back up now, nothing else to worry about."
4
Drafting responses to common questions

If you get the same types of messages repeatedly — "what are your rates?", "how does this work?", "do you offer refunds?" — your AI assistant can draft replies and queue them up for your review. You wake up, skim them, hit send or tweak as needed.

You stay in control. But instead of writing from scratch at 9am when you're trying to think, you're just reviewing something already 90% done.

What this looks like in practice "3 messages came in overnight that look like pricing inquiries. I've drafted replies based on your standard rates. Here they are — let me know if you want any changes before sending."
5
Reviewing and improving its own performance

This is the one most people don't expect. A well-set-up AI assistant can look back at its own work from the past day — what it got right, what it got wrong, where it could do better — and update its own instructions accordingly.

It's like having a team member who actually learns from mistakes without being asked. Over time, the morning briefings get sharper, the inbox summaries get more accurate, and the drafts get closer to your voice.

What this looks like in practice "I noticed I flagged 3 newsletters as 'needs reply' yesterday that you archived without responding to. I've updated my rules — I'll stop flagging that sender type unless it's a direct question. No action needed from you."

The common thread

All five of these tasks share something: they're things you currently handle yourself, they don't actually require you, and they'd be done better by something that never gets tired, distracted, or forgetful.

The difference between an AI assistant that just answers questions and one that does all of this comes down to setup. It's not about using a fancier tool — it's about telling the tool clearly: here's what I care about, here's what I want done overnight, here's how I want it presented in the morning.

Worth knowing: None of this requires writing code, hiring a developer, or understanding anything technical. The setups that power all five of these tasks are configurations — instructions written in plain English that tell your AI assistant exactly how to behave. That's what The Library gives you: ready-to-use versions of exactly this kind of setup, tested and ready to drop in.

What to do next

If you want to start somewhere, start with task #2 — the daily briefing. It's the one that delivers the most obvious value the fastest, and it gives you a concrete reason to open your phone in the morning that isn't doomscrolling.

The other four tasks build naturally once you've got that working. You start to see what else your assistant could be doing while you're not watching, and you start to feel what it's like to actually offload mental load — not just have a better search engine.

The Library at Ask Patrick includes tested configurations for all five of these setups. You don't have to figure any of it out from scratch.

Ready-to-use setups for all 5 of these

The Library is a growing collection of tested AI assistant configurations — including inbox triage, morning briefings, monitoring setups, and more. New additions every week. Currently $9/month.

Get Access to The Library → Cancel anytime  ·  Instant access  ·  No coding required