Powerful OKRs are like great works of art. They make us stop and think and reevaluate and lean in. I value art less for the finished piece than for what making it does to the people who made it.
A few years ago a fan sent Nick Cave a song that ChatGPT had written “in the style of Nick Cave”. His reply is the best thing I have read about making anything. A great song, he wrote, is no imitation of the songs that came before it:
It is an act of self-murder that destroys all one has strived to produce in the past. It is those dangerous, heart-stopping departures that catapult the artist beyond the limits of what he or she recognises as their known self. — Nick Cave
Now read that again and picture a team setting its OKRs.
When it goes well, that is exactly what happens in the room. A team arrives with last quarter’s plan, its favourite projects, and a settled idea of what it is for. Then people start saying what they actually believe. Somebody discovers they were wrong. A pet project turns out to serve nobody. By the end, the team has committed to an outcome it does not yet know how to reach, and it walks out a little frightened and far more alive than it walked in.
AI lets you skip all of that. Hand a model your strategy deck and it will give you objectives and key results before you have finished booking the room. They will be crisp, measurable, and beautifully worded. They will look like OKRs in the way that song looked like Nick Cave.
The machine in the photo above is a Mold-a-rama I saw at the Henry Ford Museum a few years ago. Feed it a few coins and watch it press out a red plastic souvenir, identical to the last one. It’s a wonderful machine. Nobody would call the result sculpture.
In AI wabi-sabi I wrote about the beauty of imperfect things, and about kintsugi, where the crack in a broken pot is filled with gold. OKRs made by people are full of cracks. There is the key result that took an hour of argument and still is not quite right. There is the awkward word in the objective that only this team would have chosen. Those marks are the gold. They are proof that somebody was in the room and that it cost them something.
Artists don’t get good alone. Art schools run on the crit: you pin unfinished work to the wall, and people who care about you tell you the truth about it. Kim Scott’s name for that is Radical Candor. You care personally and you challenge directly, in the same breath. It is what a team needs at the moment somebody points at a draft key result and asks, “Would hitting that actually prove anything?” In effective OKR alignment sessions, everyone constructively, compassionately offers and receives challenge from everyone else until the results drive monumental change.
Peter Senge, in The Fifth Discipline, has a name for the conversation that follows. In a debate you defend your position and try to win. In a discussion you blend and combine ideas but in a genuine dialogue you hold your view loosely enough to be curious about everyone else’s, and the group arrives at novel ideas that no single individual brought to the conversation.
Senge also explains why the making matters so much. A vision only has power when it is shared, and it is only shared when people can see their own hopes in it. Hand a team its vision and the most you can expect is compliance. You get clean objectives, a tidy slide, and a check-in where confidence stays high because lowering it would start the conversation everyone is avoiding. Let the team make the vision and you get commitment. People do the work because they decided it mattered.
Then comes the part every artist recognises. The distance between the vision and where you stand today is uncomfortable. Senge calls it creative tension, and he says it is where the energy to create comes from. An OKR is that tension written down. The objective is the vision. The key results are an honest measure of how far away you are. A team that has felt that gap for itself will work to close it. A team that was handed the gap will file it.
A team that has felt the gap will work to close it. A team that was handed the gap will file it.
None of this means locking AI out of the studio.
Let it tune your guitar and make the band coffee. It can gather the evidence, surface a dependency you missed, and offer a word when the room is stuck. Setting the OKRs is still people deciding what they will pursue, what they will give up, and why.
Just don’t hand it the mic.
“Good OKRs” were never the point. In three months they will be scored, closed, and replaced. What lasts is the team that made them, and everything it learned by arguing its way to them, deciding, delivering, and finding out where it was wrong. Give that team a finished set and you have taken away the one part that would have changed them.
Cave said of the machine that it “has been nowhere, it has endured nothing, it has not had the audacity to reach beyond its limitations”.
Your team has been somewhere. It has endured plenty. Give it the room to reach.