Sharick Journal

Notes on working with AI.

Comparisons, practical questions, and what to look for when an assistant starts taking initiative.

Selected Slack messages connected to a reviewed commitment and follow-up card

Slack · Follow-through · Oct 14, 2026

Slack commitment tracker: keep promises in view

A narrow workflow for capturing Slack commitments, checking the source and reviewing the next step.

A sculptural paper bridge with a gold seam connects scattered envelopes and notes to one moss-green action card with an ivory checkmark

Work · Follow-through · Oct 12, 2026

AI accountability partner for work: follow through

A useful partner remembers the promise, notices what changed and helps you choose the next move.

A compass connected to a calendar, illustrating priorities and coordination

Comparison · Oct 11, 2026

AI chief of staff vs AI executive assistant: which workflow do you need?

Do you need help seeing what matters, or carrying out a task you already chose?

Email promises organized into You owe and Waiting on cards, connected by a gold thread

Email · Follow-through · Oct 11, 2026

Email commitment tracker: keep work promises in view

What you owe, what you’re waiting for and the next action. A two-minute guide to tracking promises in your inbox.

A person considering a branching workflow and two focused work cards

Comparison · Oct 11, 2026

Lindy alternative for tracking commitments and follow-ups

Choose an assistant around the promises you keep losing.

Meeting notes becoming clear commitment cards connected by a gold thread

Follow-through · Oct 11, 2026

Meeting commitment tracker: turn decisions into follow-through

The meeting ended with a decision. Who agreed to do what next?

Two paper workflows with email, calendar and briefing cards beside a central choice

Comparison · Oct 11, 2026

Sharick vs Lindy: which fits your follow-up workflow?

Both help with follow-ups. Start with the work you want help carrying.

Two always-on agent cards, one for Microsoft Autopilot and one for OpenAI dots, each with work waiting for approval, beside one short Sharick brief that narrows a day’s handbacks to what needs you

Always-on agents · Oct 8, 2026

Microsoft Copilot Autopilot vs OpenAI dots vs Sharick: what’s the difference?

Both keep working after you log off. Where they live, whose identity they act under, how you pay, and the question both leave to you: what reaches you.

A to-do card whose date has been crossed out four times, with reminder bells piling up, beside one calm card that asks: still relevant?

How-to · Oct 6, 2026

When you keep postponing your own to-do, ask it a question instead of setting another reminder

A to-do you keep moving is a signal, not a character flaw. Three ways to read it, four honest answers to give it, and a weekly practice for the one that is hard because it matters.

Essay · Oct 2, 2026

The AI era’s operating system is a dispatcher, not an app store

System 1 models just made it affordable for AI to go first. Why that makes an AI-era operating system possible, and why its core is a dispatcher.

Three separate morning messages from ChatGPT, Claude and Gemini, next to one Sharick brief that knows what you promised

How-to · Oct 1, 2026

How to make ChatGPT, Claude and Gemini proactive (and where DIY stops)

Step-by-step setup for ChatGPT scheduled tasks, Claude Cowork and Gemini scheduled actions, three prompts to start with, and where scheduling stops being enough.

Red unread badges from WhatsApp, Slack, LINE and Telegram collapsing into one Beeper inbox, then one Sharick card with a drafted reply you send yourself

Messaging · Oct 1, 2026

How to manage messages across WhatsApp, Slack, LINE and iMessage with AI

Your chats are split across apps your AI can’t read. Get them into one inbox with Beeper, then let Sharick track what you owe and draft replies you send yourself.

Three inbox cards for personal, company and side-project accounts, each hiding one small promise, threaded into one brief that says you owe 3 replies across 3 accounts

Email · Oct 1, 2026

Managing multiple email accounts with AI: why a unified inbox isn’t enough

Gmail removes Gmailify and POP in January 2027. Your options for multiple inboxes, from the Gmail app to Spark, Superhuman and Shortwave, and why one view of what you owe matters more.

A promise from Monday, four silent days on a gold timeline, and a follow-up draft waiting for review

Follow-ups · Sep 30, 2026

How to Keep Track of Promises at Work | Sharick

Why small promises slip, the five fields behind a trustworthy follow-up, and when an assistant should nudge or stay quiet.

Gold lines from many directions converging on a single brain icon, beside the headline Your brain is.

Essay · Managing AI agents · Sep 30, 2026

AI isn’t the bottleneck anymore. Your brain is.

We scaled the workers, not the person they report to. Why running many AI agents jams your attention, and why you need an Attentive Dispatcher.

A call bot, a laptop and a pendant each capturing a meeting, threaded into one briefing that says You promised: send the revised proposal by Friday

Meeting note takers · Sep 30, 2026

AI meeting note takers in 2026: Granola, Omi, Plaud, tl;dv, Notion and Sharick

Bot, no bot, or wearable? How six note takers capture a meeting, and who helps with the promises made in it.

A blurred pile of notifications with one card in focus: Maya hasn’t replied in 3 days, the other 23 can wait

Essay · Attentive AI · Sep 30, 2026

Proactive is not enough: the case for attentive AI assistants

An assistant that goes first is not the same as one that pays attention. Why noticing absence, weighing priorities, and staying quiet matter more than speed.

Four always-on agents labeled dots, OpenClaw, Muse and Grok Bot each sending status messages toward one brief that says 3 need you and 11 can wait

Always-on agents · Sep 30, 2026

OpenAI dots vs OpenClaw, Muse and Grok Bot: always-on AI agents compared

Is OpenAI’s new dots just a worse OpenClaw? Model, hosting, price and guardrails compared, plus the question none of them answers: who manages what they hand back.

Split view: an empty Ask anything box on the reactive side, a brief about an unsent investor deck on the proactive side

Understanding proactive AI · Sep 30, 2026

Proactive vs reactive AI assistants: what changes when the assistant goes first

How the two fail differently, when reactive is enough, and four checks before you switch.

About ten agent streams flowing into a one-screen morning digest with Today, Decide and FYI sections

Field notes · Sep 30, 2026

How I run a team of AI agents as a solo founder without becoming the bottleneck

Ten agents in parallel solved my output problem and created an attention problem. Five rules for keeping what comes back to me down to what should.

Many small notification cards on the left narrowing through a gold thread into one short brief with a single highlighted item

AI chief of staff · Sep 29, 2026

Best AI chief of staff tools in 2026: which ones cut your interruptions?

Sharick, alfred_, Bond, Lindy, Nerve, and the free phone baseline, compared by what each decides is worth interrupting you for.

Paper cards showing a browser window, a chat, an approval button and a checklist, joined at the center by one gold thread

The personal agent wave · Sep 28, 2026

Personal AI agents in 2026: OpenAI dots, Muse, Grok Bot, Instinct and Sharick

Four new agents that act for you, and one that helps you decide what is worth acting on. What each does, how it is proactive, and where you stay in control.

Paper-cut illustration of email, a calendar, connected knowledge and a completed task linked by a gold thread

The first guide · Choosing an assistant · Sep 25, 2026

Proactive AI assistants: what they do and how to choose

What proactive assistance means, how Sharick, Saner.AI, Glean, and Lindy differ, and four small tests to find your fit.