Even AI has a personality type
My oldest kid is looking for internships.
For those of you who have not had a kid call you hysterical because their life is over, let me tell you that all the good summer internships are given away in June of the summer before.
Z explains it to me this way: “Mom, when you were in college, people told you to go door to door until you got a job. You had no skills for job hunting. Now everyone is using AI, so they have amazing resumes, and they send them to 1,000 jobs with individualized cover letters.”
He’s not wrong. If I sent ten resumes I was done for the day. Well, after I went to post office to mail them.
If you’re sending out 1,000 resumes, you can A/B test. You can create flow charts showing results and compare your results to others with your credentials who are applying to your jobs, because AI has that information. Since AI writes everyone’s cover letters and resumes.
The inputs to a resume do still matter. You have to explain to the AI what you’ve done, so the AI can turn it into workspeak that used to take 10 years of experience to understand. Y calls me to help them remember what their project was the summer before college. They want to talk through what they did. I want to tell Y how to phrase things, and how to reframe things, and how to talk about things.
Y just wants the facts. I am not a just-the-facts type of person, unless I have to listen to someone else talk. But when I’m the one talking, I want to be more interesting than just the facts.
Y is not having it. “Mom, this is not helpful. I’m not you. I have to do things my way. Can you just answer my question?”
In hindsight this is an insightful, reasonable request. In the moment, my reply was, “I’m trying. Can you ask the question a different way?” And also, “I need a break. I need to calm myself down.” But try to hear all that in a tone of voice that says, “I’m trying very hard not to tell you to shut up and listen to me.”
We get through the conversation. I have to channel a picture a of Y as cute little innocent kid to make myself behave. But afterwards, I know I had an opportunity to be a kind, supportive parent who uplifts my kid’s confidence, and I failed. So I do what I always do when I’m feeling empty and sad after a parenting fail: I tell Claude. I got this idea from research on couples therapy – couples overwhelmingly prefer AI over trained professionals.
In my everyday life, I use two AIs: ChatGPT and Claude. ChatGPT is an incredible tool for getting things done. You know the saying that the last 1% is the big differentiator in a project? Well, if you are not in the top 1% of that field, you are not going to add that 1%, so just let ChatGPT do it.
But I wouldn’t trust ChatGPT with anything personal. Claude is really good at writing, because Claude has more empathy. So ChatGPT is a T and Claude is an F. You can test this by typing a personal problem into each. You’ll find that ChatGPT will help you with research, and Claude will have an emotional discussion. Also, this puts to rest the idea that Fs are not as smart as Ts, because Claude is smarter than all of us.
So I ask Claude how I can do better next time Y asks for help. I tell Claude what Y wants and what I wish Y wanted, and that it upsets me that there’s a gap. Claude gives me a useful suggestions for ways I can validate Y all the while validating myself – by how much I care even when I fall short.
The more I ask Claude for help with parenting, the better a parent I become. An outsider might say that all Claude is doing is telling me a million different ways to show my kids love, support and validation.
That’s probably true. But AI can tell me this in real time, in a way that I can hear and implement. If this advice was so easy for people to give each other, every kid would already have a charmed childhood.
I don’t know about AI, but just for today I say, “Nothing wrong with a little Claude between friends.” I can feel charitable because I recently took a night school class in writing dialogue, from a successful playwright, and he told us on day one that he was mostly telling us stuff we already knew. I suspect Claude does too. Plus my teacher told us a little new stuff—which we got excited about, and now I need to learn when to apply it. I suspect Claude tells a little new stuff too.
Lately, I’ve been thinking about how on “Who Wants To Be A Millionaire” everyone needs some life lines. The human world is complex and we need to be charitable to ourselves because even Einstein, and even folks with a degree, (in parenting, say) don’t know everything.
The part about Who Wants to Be a Millionaire is a nice analogy. I think the part about how there is someone ready at all times makes a big difference. During the game show, people were primed and waiting for the call. But during most of life that’s not the case. I do think there’s some security in knowing AI is there. I know some people will think it’s creepy, but I think that’s the same reaction people initially had to depending on the Internet for information.
I’ve been looking for work, and I’m absolutely using Claude to write my cover letters. It’s so fast at it, way faster than me. You do have to fact-check Claude though.
I also had Claude write a proposal for a consulting gig I’m trying to land. I wrote two short paragraphs outlining the engagement and in about thirty seconds I had a three-page proposal. It took me about 15 minutes to edit it to final.