The panel keeps saying "human," and your team should borrow the line

Panels and comment threads in the design world keep landing on the same word: human. As AI tools get faster and cheaper, the people judging the work keep insisting that judgment and story are what matter, not speed. That framing is worth stealing before your next roadmap review, when someone asks why your team isn't shipping twice as fast with AI. Let me catch you up.

Speed is not the same as value

At a Trimble SketchUp panel in San Diego, Perkins Eastman's Omar Calderon Santiago said "fundamentally, architecture is a human endeavour" and warned that fast does not mean good. His line: "With the speed of things, it's very easy to rabbit hole yourself. I think we need to slow down and try to get back to the root of what we're trying to do."

That is a useful thing to say out loud when the pressure is to move faster. The point is not to reject the tools. It is to remember that shipping more, sooner, does not automatically mean shipping better. Your team can move quick and still lose the thread of what the work is for.

Use AI to boost the story, not write it

Calderon Santiago starts projects with hand sketches, then brings AI in to "boost the story that I'm trying to tell." AI develops his ideas. It does not generate them. That order matters. The human sets the direction, the tool speeds up the middle.

Kestrel Labs' Marissa Ritchin, a licensed architect, put the tradeoff plainly: automate the repetitive work so people get more time for the hard parts. Her code compliance platform handles the slow, rule-checking tasks. "I think an architect will always be a storyteller," she said. "That idea of lived experience will always inform the architect." For your team, that means pointing AI at the grunt work, not the judgment calls.

The human touch still closes the room

Here is the part worth remembering in a client or exec review. Calderon Santiago once chose hand drawings over polished renders for a presentation. The client connected with them, he said, "because he was seeing in these drawings something that someone had taken the time to make."

Effort reads. People can tell when a person cared and when a machine filled in the blanks. That is not sentiment, it is a real edge when you are trying to win trust. A slightly rougher thing that clearly came from a human can land harder than a perfect output nobody sweated over.

The line between tool and author is getting sharp

When Dezeen ran AI renders of Gaudí's unbuilt hotel, readers split hard. One called it "AI slop"; another said "someone typed a prompt into a chatbot... Clap... Clap... Clap..." But others pushed back, arguing the user practiced to get that quality, and maybe "craftsman" is the right word, not "artist."

That fight is coming to your product too. The useful takeaway is the distinction one commenter drew: it was made by an AI user, not an AI. Someone still owns the taste and the choices. Naming who authored the work, and holding them to it, keeps quality from drifting into anonymous output.

The deep cut

The strongest version of this argument is not "defend craft because craft is good." It is what a designboom podcast at Milan showed: regional Indian architecture stayed invisible for years not because the work was weak, but because nobody told its story. The lesson for you is that story is what carries good work to the people who decide. If your team lets AI flatten the narrative into generic output, the work can be strong and still go unseen. So when you defend craft in a review, tie it to outcomes: the human story is how the work gets understood, trusted, and picked. That is a business case, not a taste preference.

Three questions for your team

• Where in our process is AI generating ideas when it should be developing ideas we already set? Draw the line and write it down.
• Which repetitive tasks can we hand to automation this quarter so people spend more time on the hard calls?
• In our next big review, what is the human story behind the work, and who on the team owns telling it?