The AI Edition (or Eric finally starts to use ChatGPT)
|
|
This is the Too Many Trees newsletter, where I (Eric Nehrlich) share what I’ve been writing and reading in the realm of leadership and personal development. My executive coaching practice helps leaders amplify their impact by focusing their time and energy on what matters most, while uncovering and reshaping the unconscious patterns that may be holding them back. If you know somebody that could benefit from my perspective, please forward this to them or have them set up a free intro chat with me.
|
|
|
|
|
|
I have been avoiding the LLM / AI craze but I finally started using ChatGPT the last couple weeks and have actually found some use cases for it.
|
So what changed? Why did I start using ChatGPT? Because I’m bad at writing hooks. Like… really bad.
|
I was writing a LinkedIn post and couldn't figure out a hook I liked, so I pasted it into ChatGPT and asked it for help. It gave me 15 hooks in ten seconds. And most were better than mine.
Why didn’t I try this sooner? Because I’m stubborn - I take pride in my writing, and asking for help feels like cheating.
|
Plus I had gotten so used to being bad at it that I didn’t even think to get better - I thought the hooks I wrote were good enough. If people didn't read my stuff because the hook wasn't catchy, I figured it was their loss.
But I write these posts to help people, and I can't help people who don't see my posts. My stubbornness was limiting my impact.
|
Here's the important thing for me to remember: LLMs create output based on publicly available content, so they deliver above-average output for most tasks. They aren't going to be better than you at the stuff where you are an expert, but if you're not good at something, the LLM will likely be better.
|
And in my case, even though I'm good at a lot of things, ChatGPT was definitively better at writing hooks. It also gave me some great tips to simplify my posts and make them easier to read. Using ChatGPT will let me focus on my strengths, while increasing my impact by improving on my weaknesses.
In other words, I had to stop thinking of asking for help as cheating; instead, it lets me have greater impact and help more people.
Where might you be limiting your impact because you don't want to ask for help? Where have you been thinking you're good enough, while not admitting that you're below average and could use even the average output of AI?
|
|
|
|
Here are a few other tasks I've given ChatGPT over the past couple weeks since deciding it could be helpful:
|
- I described my ideal coaching client, and asked how I could improve my webpage to appeal more directly to that person. I haven't made the changes yet, but it had a number of good suggestions, both in terms of targeting that person more specifically, and in organizing my page more cleanly (which will also probably make it more easily scannable by ChatGPT to get me more clients).
- I asked it for date ideas for an outing with my wife. While the ideas it came up with were generic, it was still helpful to expand my thinking a little bit.
- I asked it for investment advice on what I could do if I wanted to be a little less conservative in my investments (yes, I'm the boring person who buys index funds and holds them). I won't give it control of my portfolio, but found it helpful to get ideas to start exploring other options.
What I realized is that using ChatGPT is not giving away my judgment, or thinking that it knows better than me. But it can provide a different perspective, since I can get lost in my own thinking especially as a solopreneur, and offer me a wider range of options beyond what I already know.
|
And I've seen how great it is for others who benefit greatly from getting access to above-average output on demand, from immigrants whose written communication improved greatly, to entrepreneurs able to quickly gain expertise in functions like marketing and finance and copywriting. There are positive possibilities here.
|
|
I'm also aware of the many problems with using LLMs: they have stolen and copied the work of people at scale without credit or compensation, their consumption of power and water may lead to environmental catastrophe, and they exacerbate existing power and privilege differentials. They are colonial capitalist tools that extract knowledge from others and resources from the environment and profit from that extraction. Can I justify supporting these corporations in their rapacious desire to control the world? Perhaps not. So like most technologies, there is value and there are drawbacks - I'm still deciding in which cases the value exceeds the drawbacks.
|
|
|
|
|
Looking for new coaching clients:
|
Many rising leaders reach a point where the habits that once fueled their growth (like overdelivering on every problem and request) start holding them back. Progress requires learning to let go of those reflexes, and practicing new behaviors that initially feel awkward and uncomfortable.
|
That’s where coaching can help. I work with high-performing tech executives to accelerate through this shift so they can lead with clarity, purpose, and balance instead of burnout.
|
|
|
|
And now for the normal personal development content…
|
LinkedIn: These are ideas that have helped my clients (or myself), and that I share via LinkedIn to help a wider audience, and archive here.
|
- Many of us stay stuck in situations that drain us because we forget that we’re allowed to choose differently. I share a story of going camping with my kids, having a miserable sleepless night, and then almost doing it again because I felt trapped by arbitrary self-imposed rules that I must keep my commitments and that quitting equals failure. This post was to promote my book, You Have A Choice, that was on sale during Prime Days last week.
-
|
Your resume is an advertisement, a tease to get the reader interested enough to reach out and learn more about you. Every great advertisement is essentially a story of how a person had a hole in their life, but then found a product which filled that hole and led to a happily ever after life. You are the product in this advertisement, so tell the story of the hole that you can fill for the hiring manager.
|
- I finally watched KPop Demon Hunters on Netflix and found it surprisingly relevant to my work as a coach. I share how the story can be interpreted an exploration of trauma, when somebody told us we weren't good enough and we internalized their words. And we keep replaying those voices (our "demons") believing that we can never change or grow because that is who we are. But we can escape the negative loop through sharing and compassion.
- Most people aren’t stupid or evil. They just live in a different world with different assumptions than you do. If someone’s behavior doesn’t make sense, it usually means you’re missing a key belief, incentive, or context that would make it logical in their world. I like to ask myself: “What would have to be true in their world (but not in mine) that would make their position obvious?” When I find that missing piece, I can communicate with empathy and respect, and people are far more likely to listen, because I have taken the time to understand their world.
Here are a few other frameworks for how AI can best be used:
|
- How to Use AI Without Becoming Stupid by Cedric Chin of Commoncog. The rule he proposes (via Vaughan Tan) is: "Do NOT outsource your subjective value judgments to an AI, unless you have a good reason to, in which case make sure the reason is explicitly stated." In other words, AI can't decide whether one thing is better than another or worth doing, because it doesn't have its own set of values to decide what's most important. It's good for exploration (learning about options you might not know about) or for upgrading your capabilities in some area (as I did with writing hooks), but should not be used for life advice or anything that requires it to make value judgments.
- Duncan Brown shares a talk about Team dynamics after AI. He shares his experience that AI accelerates the part of the process that isn't the bottleneck. It makes it quicker to generate code and prototypes, but the hard part of software development is getting agreement on what to build: "being able to get feedback on your artefacts is much, much more important than the artefacts themselves". By flattening out diversity of perspectives (letting the AI's generic judgment guide design choices), by bypassing translation between different perspectives (learning to actually communicate across our differences), and by limiting feedback into the system (so that it can learn from when it falls short of our expectations), AI may be making things faster but is losing relevance to our particular problems.
- Henry Farrell makes a similar point that The Map is Eating the Territory, in which he suggests that LLMs are not intelligent per se (since they rely on human generated knowledge), but instead are "cultural technologies" (a phrase coined by Alison Gopnik) that "provide new ways to access, order and remix human generated information". He compares them to markets and bureaucracy, both of which are methods by which humans can coordinate their actions at far greater scale than was previously possible. But such technologies create problems as well as possibilities; the one he foresees for LLMs is what is alluded to in the article title, that "LLMs are cultural technologies of summarization, whose value depends on people continuing to actually produce culture that can usefully be summarized. [but] the summarizations that they produce risk supplanting the culture that they feed on."
|
|
Thanks for reading, and see you in a couple weeks!
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
