03

The meeting you attend in 2030 will look nothing like the one you're dreading on Monday

Some things will disappear. Some will become more important. The question is whether you're designing for what remains.

The business meeting is approximately 5,000 years old.

The Roman Forum was the most celebrated meeting place in the ancient world. Medieval guilds held formal gatherings every six months. The etymological root of "company" is the Latin companio — to break bread together. Humans have been sitting in rooms talking about work for as long as there has been work to talk about.

This matters because every generation believes it has discovered the meeting problem. Every generation has been partially right. And every generation's solutions have, sooner or later, produced more meetings.

AI is the latest candidate for meeting salvation. Here is what I think it will and will not change.

The meetings that will largely disappear

Will disappear
The Status Meeting

"Just checking in" meetings exist to signal availability and signal work. AI agents can handle status reporting: monitoring project state, surfacing anomalies, alerting when a human decision is needed. High-frequency, low-information — a poor use of human time.

Will disappear
The Briefing Meeting

Meetings that exist to share information that could have been a document, an email, or a short recorded video will increasingly be replaced by exactly those things. By 2027, AI agents will attend on your behalf, summarise, flag what requires your attention, and draft your response.

Will largely disappear
The Pre-Prep Meeting

Prep meetings exist because people arrive unprepared. AI changes this. Pre-meeting briefings — background on attendees, summaries of relevant context, draft agendas with questions — will be generated automatically. No need to meet to prepare to meet.

The meetings that will remain — and matter more

Will remain
The Trust-Building Meeting

You cannot outsource the process of deciding whether you trust someone to an AI agent. Trust is built through shared experience, through watching how someone behaves under pressure, through small signals no transcript captures. The first conversation, a difficult negotiation, a moment of repair after something went wrong.

Will remain
The Creative Meeting

The moment in a good brainstorm when something unexpected is said and three other people immediately see something they didn't see before — that remains hard to replicate. AI can generate options. It cannot have the genuine surprise of encountering another mind that sees differently to you.

Will remain
The Decision-Accountability Meeting

When the decision is genuinely difficult — where reasonable people disagree, where there are real stakes, where someone needs to own the outcome — the meeting still matters. A meeting makes the decision visible. It creates a shared memory that a document alone doesn't.

The ratio

Right now, the average knowledge worker attends 13.6 meetings a week. Most of those are the first category — status checks, information transfers, prep rituals. AI will handle a significant portion of that category within five years. The meetings that remain will be a smaller number, higher stakes, and more demanding.

"The meeting you cannot skip will be harder than the twenty meetings you used to sit through half-present."
The honest consequence of calendar clearing

This is good news and bad news. Good, because the calendar clears. Bad, because every meeting left standing will require full attention, real preparation, and genuine presence. You will not be able to coast.

Now
Copilot and Claude Cowork handle recap, follow-up drafting, pre-meeting briefing, and note synthesis. The meeting itself is unchanged. Most people don't use any of it.
2027
AI agents attend recurring status meetings on your behalf. Briefing meetings become async by default. Pre-meeting prep is automatic. Meeting volume begins to fall meaningfully.
2030
The average knowledge worker attends 4–6 meetings a week, down from 13.6. Each remaining meeting is trust, creativity, or high-stakes decision. Each requires more than today's meetings demanded.

Not "fewer meetings" — but "what are we gathering humans for?"

Organisations spend considerable time thinking about how many meetings to have and how long they should be. Very few spend time thinking about what meetings are now for, given what AI can handle instead.

The right question for 2025

Not "how do we reduce meetings?" but "given that AI can now handle information transfer, status reporting, and coordination at scale — what are we gathering humans for?"

The answer is probably: trust, creativity, accountability, and the decisions too complex for any agent to make alone.

If that's right, then the meeting of 2030 looks less like a briefing and more like a deliberation. Less a tool for coordination and more a space for judgment.

That is a higher bar. It requires better preparation, clearer purpose, and more honest participation than most of today's meetings demand.

The organisations that start designing for that now will not just have fewer meetings. They will have better ones.

Where to go next

Booqed
The history of meetings from agora to Zoom

Useful context on what meetings have always been for — and why every generation thinks it invented the problem.

Microsoft Research
AI is driving rapid change, uneven benefits

The best current research on how AI is changing knowledge work — rigorously sourced, not a sales piece.

MorningMate
The future of meetings: embrace asynchronous work

The case for async-first, clearly argued — with the limits acknowledged.

What do you think will be the last type of meeting to disappear?

Originally published on LinkedIn. Craig Stanley works independently on AI adoption, future of work, and learning design.


← Prev: What AI can do now All issues →