Alignment Is Eating Itself. What Comes After the Deck.

TL;DR

Most corporate alignment work (status decks, cross-functional reviews, steering committees, doc relay between layers) existed to move information through orgs that couldn't move it any other way. Agents move it instantly, and the scaffolding is collapsing.
What's left when the scaffolding falls is the thing the scaffolding obscured: actual disagreement about what to build. The new scarce skill is productive conviction under contested ground truth, not consensus-building.
Leaders who spent a decade mastering alignment theatre now need a skill they never developed. Many won't make the transition. The org will move anyway.

I've sat in thousands of hours of alignment meetings across fifteen years. Quarterly reviews. Product council. Steering committees. Pre-reads for the pre-read. Stakeholder alignment huddles. Post-meeting recap docs that existed to ensure everyone who missed the meeting understood what had been aligned on. Each of those artefacts had a reason to exist, and most of them did real work. They weren't theatre in the pejorative sense. They were load-bearing scaffolding in an operating model where information moved through human relays because there was no other way to move it.

That scaffolding is collapsing, and most senior operators haven't noticed yet because the meetings are still on their calendars.

Agents don't need to be aligned; they read ground truth directly. Documents route themselves. Status reports generate from the underlying systems. Cross-functional briefings are produced in thirty seconds from whatever source material you point at. The mechanical work of alignment, the stuff that was 60–80% of what senior operators actually did all day, is being absorbed by the systems. What's left is the part that was never about alignment at all: actually disagreeing about what to build, and deciding.

What the scaffolding was actually doing

Alignment rituals performed three distinct functions, which got conflated.

Information transport. Moving facts, updates, and state between people who needed them. The PM told the VP; the VP told the exec; the exec told the board. Each transport hop required a meeting, a doc, or a ritual because the information couldn't move another way. At each hop, some signal was lost, some spin was added, and the cycle time was measured in weeks.

Decision confirmation. Getting enough people to say "yes" to a proposal that it couldn't be reversed without embarrassing all the yes-sayers. Sometimes this was genuine conviction. More often, it was risk-distribution: if the decision turned out badly, the responsibility was spread across everyone who had aligned on it. The ritual was the thing that made the risk-sharing visible and binding.

Political theatre. Rituals whose purpose was to demonstrate who had influence, whose priorities were important, and whose team was ascendant or struggling. Most senior operators could read the political subtext of a status meeting in ten seconds, because the subtext was the point.

The first two functions are the ones AI erodes. The third is the one AI exposes.

When information moves instantly and cleanly, you don't need rituals to transport it. When decisions can be made against ground truth by people who can see the ground truth directly, you don't need rituals to confirm them. What remains is the political subtext, which used to be embedded in otherwise-useful meetings and now has to stand on its own. A lot of what used to pass for alignment work is becoming visibly naked when the information-transport and decision-confirmation layers are stripped away.

The two skills alignment theatre was hiding

The uncomfortable implication sits with senior operators who built their reputations on running alignment well. Most of them did not, along the way, develop two skills they now need.

Productive disagreement. The ability to disagree hard, in writing, in front of other senior people, with ground truth, with specific evidence. Alignment theatre rewarded the softer version of disagreement: the tactful question, the concerned raise, the suggestion offered diplomatically, and the 1:1 backchannel. Those skills still have a place. They are not sufficient. When the meeting's purpose is not consensus but decision, the people who can argue cleanly and directly carry most of the load. The people who can only disagree through gentle implication get quietly talked past.

Conviction under contested ground truth. The willingness to commit to a view when other senior people are committing to different views, backed by the same data. Alignment theatre used to absorb contested views by running them through enough rituals that something merged out the other side. When the rituals are compressed, the contested views have to resolve in hours or days, and they resolve through conviction, not through process. Leaders who built careers on never being clearly wrong about anything suddenly find themselves in a context where being clearly right (or clearly wrong) is the daily requirement.

Neither of these skills was strictly optional in the old model, but both were substitutable: you could have an alignment-heavy career without ever having to do them well if you were skilled enough at the other thing. In the new model, the substitution path is disappearing. The operators who thrive in an agentic, flatter operating model are the ones who can actually decide, hold conviction, and take the political cost of being wrong.

What replaces the alignment meeting

The first question senior operators ask me when I describe this is "what does my calendar look like in the new model?" Fair question. The answer is less dramatic than the framing suggests.

Three patterns emerge:

Shorter, harder meetings. The quarterly product review shrinks from two hours with thirty slides to thirty minutes with a single prompt fed to an agent that ingested the last quarter of product metrics, customer feedback, and eval data. The meeting is no longer about reviewing the status; the status is already common knowledge. The meeting is about the one hard decision that the status surfaced. The hard decision is the only thing worth a room of senior people's time.

Async conviction over synchronous consensus. A lot of what used to be "we need to get everyone in a room" becomes "we need to get everyone's written position before we meet." Written positions, backed by evidence that all participants can see, compress debate time enormously. The meeting then happens to resolve the actual disagreements, not to surface them. Senior operators who've been promoted through their talent for in-person persuasion find that this inverts the advantage; senior operators who write clearly find that it amplifies it.

Fewer standing meetings, more situational ones. Most standing alignment meetings existed because information had to flow on a predictable cadence. When information flows continuously, the meetings can be called when something actually needs to be decided. Many of them don't need to be called at all. Standing meetings are the single easiest thing for a new operating model to prune; most orgs don't do it fast enough, and the calendar drag alone slows the transition by quarters.

Managing this transition is an org-design problem, not a tool-selection problem. The teams that get it right don't add AI tools to existing meetings; they rebuild the meeting layer around what agents now do, and aggressively cut what doesn't survive the rebuild.

The resistance pattern

The org will push back on this. Predictably.

Senior leaders will defend the meetings they run. They're not being stupid; their status and influence are bound up in those meetings, and cutting the meeting feels like cutting the authority it represents. The quarterly review is a visible artefact of their authority. Cutting it feels like a diminishment. I've watched VPs argue strenuously for the preservation of rituals that no one below them found useful, because the ritual was the thing that made them visible.

Middle management will defend the alignment chains. The PM who ran five alignment meetings a week was producing real outputs for each of them (the prep doc, the pre-read, the summary, the follow-up). When those meetings are cut, those outputs don't have a home. The PM feels, correctly, that their value is tied to the outputs, and resists the cut that makes the outputs obsolete. The fix is to change what the PM is accountable for, not to preserve the scaffolding.

Junior operators will accept the change fastest. They didn't build their careers on alignment theatre. They didn't have much of it to begin with. When senior layers cut the rituals, junior operators find themselves suddenly freed from prep work they hated, with more time for substantive work. They're the easiest population to bring along, which is why successful transitions often look like they came from the bottom up even when the strategic call was made at the top.

The path that works: cut the rituals the organisation doesn't depend on first, rebuild the decision paths that remain around async conviction, and accept that 20–30% of senior operator time will genuinely need to be redirected from ritual-running to actual decisions. That reallocation is the real work. It's harder than installing an AI tool. It has better long-term returns.

What senior operators should do

If you're reading this and uncomfortably recognising your own calendar, three practical moves:

Audit your standing meetings against a simple test. For each one: if I killed this meeting tomorrow, would anything break in the next 30 days? If the answer is "not really," kill it. If the answer is "my team would feel disconnected," that's a human problem, not a meeting problem, and the meeting isn't solving it. Most senior calendars have 3–5 meetings that fail this test today.

Write your next major decision down before the meeting. Not a deck. A two-page argument with ground truth and a recommendation. Share it before the meeting. Ask participants to do the same. See what happens to the meeting. In my experience, the meeting either becomes dramatically shorter and sharper or becomes unnecessary. Both outcomes are improvements.

Practice disagreement. Out loud. Pick one decision this quarter where you genuinely disagree with a peer and write your position clearly, including why they might be right. Share it. The muscle of productive disagreement atrophies fast if you only exercise it through polite implication, and you need it more now than you did three years ago.

None of these are revolutionary changes. They're calibrations. The cumulative effect of running them deliberately for a quarter is what separates the senior operators thriving in the new model from the ones wondering why they're being quietly routed around in their own organisations.

The CMO signal

One of the most useful signals about where this is going: the CMO is often the heaviest token consumer in an AI-adopting company, ahead of engineering, because they're producing direct work product with AI rather than managing people who produce work product. That's the senior-operator profile that survives the transition: people who use the tools themselves, make calls with the outputs, and bear accountability for the decisions that result. The senior operators who don't get there become the barrel problem their org is exposing in real time.

Alignment theatre isn't dying because AI replaced human judgment. It's dying because the scaffolding around human judgment is being compressed, and what's left is the judgment itself. That's a better operating model. It's also a harder one for leaders who trained on the scaffolding.

The senior operators who thrive in the next five years won't be the ones who mastered a new tool stack. They'll be the ones who mastered a new decision posture: shorter, sharper, more accountable, more willing to be clearly wrong about something today because the cost of being clearly nothing about everything is higher.

Frequently Asked Questions

Are you saying alignment meetings are all useless?

No. Some alignment meetings still do real work: making decisions that genuinely need several senior people in the room at the same time, resolving honest ambiguity about priorities, or building the shared context that new members of a team need. The argument is that these decision-making meetings were 20–40% of the alignment calendar; the other 60–80% was information-transport that agents now handle. Cutting the latter doesn't touch the former.

What about alignment up the chain, to the board or investors?

Same dynamics, slower timeline. Board decks and investor updates used to be high-effort synthesis exercises. They're increasingly generated from the underlying systems with senior operators editing for narrative and emphasis rather than writing from scratch. The meeting with the board will still happen; what gets discussed in it is already changing, and the discussion is getting sharper because everyone has seen the ground truth before the meeting rather than in it.

How do I run this transition if my boss is still in alignment-theatre mode?

Carefully. You can't unilaterally cut the meetings your boss values. What you can do: produce async written positions in advance of those meetings, so the meetings become sharper. Offer to run shorter versions of standing meetings as pilots. Let the quality-of-decision difference do the argument for you. Senior leaders usually come around to this model faster than you'd expect when they see a version of it working in their org.

Doesn't sharper disagreement just mean more conflict?

Less conflict, typically, because the conflict is surfaced earlier and resolved once instead of festering across three quarters of tactful alignment meetings. The instinct that disagreement equals conflict is itself an alignment-era artefact. In operating models that reward clean conviction, people who can disagree well are trusted more, not less. The relationships are stronger on the other side of a sharp disagreement than they were on either side of a decade of tactful avoidance.