The Leaflet #123: building identity on day -1; a thought experiment using younger you; AI as ClipArt
Building identity and agreement on day -1
There’s a kind of agreement you can arrange on day one that you can almost never arrange on day one hundred. One of the most important of these agreements is that the person joining your team is hungry for direct, corrective feedback. The way they show that hunger is by acting on the feedback without delay, friction, residue, weirdness, other faff.
You can ask someone to commit to that during hiring and they’ll mean it. If you ask them to commit to it months and months in, it’s harder to get what you want and need. After several months, you have a relationship that already has its own grammar: what kind of comment means what, what tone signals what, what’s said and what isn’t. You can refine that grammar but replacing it is tough.
So in a fast-growing org, where you’ll run this loop dozens or even hundreds of times in a few years, the hiring process is a durable lever for setting the subtext you’ll rely on for years.
I suggest two moves to get this agreement set:
First, test for it. In a late-stage round, give the candidate a real piece of direct feedback on something they did during the process. Watch what they do with it. Do they act on it before the next exchange? Do they thank you without sounding wounded? Or do they explain themselves at length, dispute the premise, emit anti-social vibes?
Second, model it. Bring the candidate into a real one-on-one where you give an existing teammate — someone you’ve picked because they receive feedback beautifully — a piece of pretty direct feedback. Then debrief with the candidate: that’s what this looks like here. If you can picture yourself in that seat enjoying that exchange, this place will work for you. If that looked grim, this won’t.
With this two-step, you’ve selected for feedback responsiveness and the candidate has chosen the identity of someone with that trait. If you skip this agreement, it’s hard to implant or backfill it. You can end up with a team that’s superficially polite, structurally defensive, and very challenging to coach.
-ben
A thought experiment using younger you
If your thinking feels stale and you’re looking for a fresh approach, run this thought experiment: how would younger you have solved this problem?
Some features of younger me that might also be true of younger you:
Thinner network
More physical energy
Less context (aka knowledge of how things go wrong and which solutions have tended to work)
Less cash on hand (harder to buy my way out of a problem or outsource it)
These cash out as a self-reliant naivete that can be really useful for prototyping an idea, just making the thing your own damn self, [fooling] around and finding out. If I’ve accumulated lots of reps in a domain over time, I can lose touch with those frisky, improvisational moves, trapping myself in an optimization game where I do the same thing over and over because I’m beholden to some narrow idea of what’s efficient, or smart, or best practice.
This thought experiment can sit in your toolkit alongside a few others we’ve recommended over time, namely:
-eric
AI and ClipArt: a millennial tale of taste
Longtime readers here know that we’re avid LLM users. We often recommend prompts and use cases in this very newsletter (for coaching, hiring, career transitions, distinguishing good from excellent, etc).
This short post is more observation and lightly held prediction than advice. LLMs can generate images, prose, and infographics now that sit between serviceable and strikingly good.1 It’s also true that much of this generated stuff carries tells. An AI-generated artifact bears signatures, subtle and obvious, of its origin. (For humorous/snarky roundups of these, check out threads like this one on … Threads).
This moment in AI usage reminds me a little of the mid-90s era of Clip Art and Word Art. Thirty years ago, you could tell when someone had quickly pulled a preset image or template off the first page of the menu in Microsoft Word and dropped it right into their document. Usually this looked better than what the same person could have made in Microsoft Paint (very limited toolkit) or Photoshop (very steep learning curve). To a trained eye or a pro designer, though, these artifacts were amateur, slapdash, even offensive. Eventually, even for the average observer, there was a flat sameness to these artifacts. The style became cliche.
As more and more millions of people start making stuff with frontier models, the default outputs of the models will become familiar and flat and cliche. In some cases, that’s just fine. You don’t need the text of an MOU to wow you with its idiosyncratic artistic flair. But if you are looking to stand out, remember that billions of other people have free access to the same insane mega-computer that you do. Your one-shot prompt will yield something that looks “better” than what you might draw free-hand. But it might also look better in the exact same way everyone else’s does.
-eric
*This entire post a shaky attempt to avoid a rant about the default GPT typeface.
COMPELLING QUOTATIONS
Poet and lawyer Reginald Dwayne Betts on time in prison:
I have called, in my wasted youth, the concrete slabs
Of prison home. Awakened to guards keeping tabs
On my breath. Bartered with every kind of madness,
The state’s mandatory minimums and my own callus.
I’ve never called a man father; & while sleep, twice
Wrecked cars; drank whisky straight; nothing suffices –
I fell in love with sons I wouldn’t give my name.
95-year-old John McPhee sharing 80-year-old memories of cream:
I was fifteen years old and always hungry. I lifted the heavy mushroom-shaped lid off a milk can and dipped a ladle into the cream. If you have ever tasted heavy cream, this was heavier. Call it viscous. Call it cream that would stand up like mayonnaise, cream that no taste bud would ever forget. It didn’t go into the cereal like milk, it went onto the cereal as a layer.
Mississippian Wright Thompson on New Orleans in 2015:
Rebirth has been the standing field order of the past ten years in New Orleans, a powerful force shaping the city in ways big and small. Everything is governed by this spirit of renewal, and everything is viewed through its lens, from the fervent love of brass bands to the New Orleans Saints, the standard-bearers of a city struggling back to its feet. But within this hopeful word an idea hides in plain sight: For something to be reborn, it must have first died.
In addition to, you know, cracking mathematics problems that have stumped leading mathematicians for years.
 | A guest post by
|