I’ve been thinking about this quote a lot lately:
“There is no compression algorithm for experience.”
It’s one that came up often during my time at Amazon. Andy Jassy, then CEO of AWS, first said it in 2017.
Waaaay back in the before times—and by that I mean pre-AI.
Since then, Jassy has become Amazon's CEO and I now work at the intersection of AI and education. But the sentiment feels more relevant than ever.
To AI or not to AI …
Is not the question.
That ship has sailed. AI has been adopted faster than any major technology in recent memory, reaching roughly 39% adoption in just two years.1
Compare that to roughly 5 years for the Internet and 12 years for PCs.
Businesses are betting heavily on AI as a driver of efficiency and growth. You can see it everywhere: job postings now list AI proficiency as a required skill, earnings reports tout AI initiatives, and CEOs blame AI for the latest round of layoffs—because AI is coming for all our jobs, apparently. 🙄
Okay, boomer.
I’m not worried that AI is coming for my job.
It’s true that AI can produce go-to-market plans, sales playbooks, prospecting emails, QBR templates—and pretty much any other asset I regularly produce or use in my role—in a fraction of the time it takes me to do the same.
Some of it is even pretty good … so I use it to support my work.
But not all of it is good. Or right. The problem is AI doesn’t know which is which. (See the recent REI bicycle ad as the latest example of this.)
It doesn’t know my audience. It can’t understand university faculty, staff, and administrators as the wonderfully complicated humans they are. It doesn’t understand institutional politics, competing incentives, or decades of accumulated context.
It doesn’t know what motivates them or how to connect with them.
That requires judgment—the kind that comes from observation, trial and error, conferring with peers, and years spent working in and around this industry.
You have to know when to trust AI, when to ignore it, and how to shape its output into something that actually works.
That takes experience. And there is no compression algorithm for that.
In the weeks to come, I’ll be exploring what this means for us as leaders, for our educational institutions, and for the edtech companies that serve them.
In the meantime, I want to hear what you think:
How does AI reshape our work as leaders, educators, and technology providers? And how do we gain experience and develop judgment in a world where AI is doing more of the work?
Hit reply to drop me a quick note or leave a comment below.
Enjoying this blog? SGNR is 100% free and always written by a human—me. Subscribe to get new posts delivered directly to your inbox. Thanks for supporting my work!
2024. Federal Reserve Bank of St. Louis, The Rapid Adoption of Generative AI.
Hi Rae, I completely agree, You have to know when to trust AI and also when to ignore it. I have seen so much rubbish generated from AI, and I have also saved countless hours using it too. One thing is still very clear: it cant sit in front of a customer and understand their pain point. It cant react to the smiles or see the weight get lifted of the customers shoulders when they feel heard. As JD says below to. It can't make you a good manager, it can't make you a good product manager or sales person, but it can really help support you!
At Davidson we pride ourselves on the human to human connection being the difference maker, the value proposition, the reason a student (or someone) pays the price of admission and all of us allegedly come to work every day. So I use AI to get the tedium and the minutia out of the way to return my focus there, and to the college's overall mission.
To be specific, AI helps me prepare for 1 on 1s by scanning email, Slack, Zendesk, Oracle goals, Github, and more... to create a complete picture of what my staff have been working on since we last spoke. It automatically creates check-ins in the HR system along with surprisingly accurate suggested topics for discussion. When meeting on Zoom with a transcript enabled, a one-sentence request to AI loads a summary of that same transcript back into the check-in, augmenting the suggested topics with what was actually discussed. For in person meetings, it even helped me write a free, open source app that handles transcription, diarization, summarization, and sync between my iPhone and my Mac.
None of these things actually make me a good manager. That's where training and experience come into play as you rightly point out. That experience came from years in the seat, trainings like the MOR Leaders Program, and the caravan of coaches and PD that have processed through the office over the last decade. But training and experience cannot (alone) save you from drowning in tedium. And so AI helps me rise above the fray a bit so I can actually focus on the person in front of me.
This is just one small example. I've used it to fill out capital budget requests (all the context was already in my email and Slack convos anyway), compare hundreds of pages of AV bid documents, create pitch decks and videos, and so much more.
How do we gain more experience? Well, I've gained enormous experience with AI these last two years in particular. And as it multiplies my impact, I can spread that impact further out into the institution, and learn about (and make a difference in) all the corners where the mission is actually happening.