Letter from the future (Because I already did 2+ years of using generative AI as work support, so you don't have to)

Hi. Yes, you. I'm writing from the future. 

I'm writing from after two years of using AI to summarize, generate, and synthesize information for (writing/desk) work. 

I'm here to say.  It's going to make you dumber.

Sorry, that was harsh.  More accurately, it's going to make you a weaker thinker.

It's great, too, of course.

You're doing it because you need to juggle a full load of school WHILE working at an early-stage start-up double-hatting while the product manager is on leave. 

It does actually seem like a situation where AI is necessary to juggle both workloads of information consumption and documentation.

You use it to make reviewers.

You use it to explain machine learning formulas to you like you're 5.

You use it to generate project plans, when you don't have time to type one up from scratch.

You use it to summarize a video transcript or research paper, so you can read the key points for a school presentation.

You even try using it to make a customer journey map. It sucks at that, but you tried.

You used it to help you write code for your AI class on coding a decision tree.

You're not basic – you don't ask it for answers, like a Google Search.  You're also not wasteful – you don't use it to make memes. 

You use it...as needed. And you need it.  So you use it a lot.

How else are you going to remember the methods for management accounting, linear programming, building a website using Razor Pages, learning PowerBI, while doing QA work for your tech job and also trying to pivot a design team to making interfaces for agentic AI? (All while doing school drop-offs, pick-ups, laundry, and preparing food for me and my kid's meals)

Well, after two years - of automatically summarized meeting minutes, of attempting to feed it qualitative/ interview data and ask it to do an initial synthesis, of asking it to draft documents, generate project plans, match lots of resume versions to job descriptions – now, it's harder for you to write.

There was even a time...that you thought your own writing started sounding like ChatGPT. 

Claude freaked you out though, it really sounded like a person.

Well, what you should know is, actually – we stop using LLMs, you and I.  Or we use it way less than you did.

We realize that if we keep using it, we lose the "muscle" of structuring information ourselves.

We realize we start getting slower/worse at summarizing and synthesizing, when we need to do it on our own.

And, that sounds small...but, we can still feel the difference in how we write, even now.

It was also hard to find the focused time and energy to write after becoming a parent, then this attraction of "just using ChatGPT" didn't help either. Having something churn out words, instead of them coming from your hands was so temptingly easy, it was eventually such a slog "waiting" for them to come out letter by letter like this.

But, now, we know what it's like when we don't try to keep at it. So we do.