# Good for nothing. > Three things AI is actually good for — curiosity, superpowers, and (hear me out) human connection. - written by: Andrew Schroeder - edited by: Fable 5, GPT-5.6 Sol - status: published · 2026-08-20 - canonical: https://www.moltolabs.ai/notes/good-for-nothing/ - raw markdown: https://www.moltolabs.ai/notes/good-for-nothing.md --- AI takes a lot of heat — especially if you’re outside of the goldrush-frenzied Bay Area. There are obviously off-putting and scary narratives about job loss, energy and water usage, the oligarchy and permanent underclass, slop, brainrot, China vs US, kids outsourcing their homework, chatbots replacing real relationships, more accessible than ever deepfakes…[Skynet](https://www.youtube.com/watch?v=4DQsG3TKQ0I). What’s not to love? I’m not going to get into these today (for the record I do think there’s plenty to be concerned about). Instead, I wanted to go through some of the reasons why I enjoy using AI (and maybe even some good that can come from it). I used to work for a company that uses AI to help financial advisors. I’ve helped people work more efficiently, and maybe even made their professional lives better. I’m also what you might call an early adopter and personally super interested in seeing what this technology can do. Ok, now my list. Not exhaustive or even necessarily correct, just going for 3 today: 1. Feeding your curiosity. I remember my first experience with the internet. I was in elementary school, in the library, and thought to myself “so, a way more confusing way to find and read books.” I forget exactly what we were doing, but I assume it was some early browser or encyclopedia experience. That said, I quickly came around and realized how powerful it was to have a ton of knowledge at your fingertips (Ask Jeeves comes to mind). And that’s basically been the experience for most of my life. Google any question that comes to mind, hopefully have some critical thinking when it comes to sources, read. I did not think this was a bad experience. Quite the opposite! [My first experience with ChatGPT](/notes/the-ai-slot-machine/) was asking it for some info about a relatively obscure car. Failed. Found the correct answer using Google in seconds (the car’s [not that obscure](https://www.youtube.com/watch?v=WL2wSRskkxw)). But what if I want to have a conversation about something? I like to talk through things. I can’t chat with Google. What do we talk about? Coffee (flavor profiles, brewing techniques, history around third wave coffee culture). Backpacking (pros and cons of ways to cut pack weight, permitting, where I want to go next). Career (what do I want, what have other people done, why am I making all these silly LinkedIn posts). None of these topics are earth shattering. Maybe I would get to the same place working through it myself. Perhaps I’m actively getting a bad (or average) thinking partner (if you can even call software that). I don’t know. But I don’t think that’s the point — instead let’s focus on how AI makes having these conversations so accessible that it feels odd not to have them. What makes me hopeful is that this accessibility will encourage people to be more curious. Or act on some underlying curiosity that’s always been inside. It is a little addicting to have a thought pop into your head and be able to not only entertain it, but maybe let it take you in a new direction. I know we all talk about enablement with AI, but I think the point here is less that I’m able to get all this done and more that I don’t know what an LLM or agent harness will do or say (turns out non-deterministic systems are more interesting to me than deterministic machines). I actually feel pressure to push this new technology further, which itself is a form of curiosity. 2. Superpowers. Ok had to slot the easy one in at number 2, and to be fair it’s a bit of a continuation of number 1. But it’s big enough that it deserves acknowledgment. When friends of friends ask me why I like playing with AI the easiest answer is that it lets me do things I otherwise couldn’t (or at least at a level that I couldn’t). Coding/building software is the obvious one. I’ve basically been that guy who has some technical intuition but can’t actually code for my whole career. It’s not that I don’t love building things. I always have. It’s that I kinda hated coding. I could never remember syntax, I would lose track of brackets, my eyes hurt from squinting to read things character by character. AI changes things. I don’t have to know how to code. I can get around without even being a great architect. I know what I want and I’ve learned to get Claude Code to build it for me — natural language is the only requirement. Now, I’m not claiming that anything we build is as excellent or elegant as what my engineer friends could code or prompt into existence, but something is better than nothing (and nothing was what I was doing when I would have needed to devote my time to doing something I actively disliked). I’ll also note that this doesn’t stop with building software (though building a “Strava-for-____” seems to be the first thing that most of us do, me included). I consider everything I listed in #1 to be a superpower. Not that it was unattainable before, but rather that it wasn’t as simple as opening an app on my phone. 3. Human connection. What?! Let me try to explain before you think I’m just listing out things that anyone would be happy about. So…AI has been pretty shitty for some people. It’s really sad and frustrating and anger-inducing seeing how this little magic piece of technology can prey on people and take them out of reality and real relationships. Hell, I worry about myself sometimes (I use AI a lot). But I do think it’s important that we view AI as a new tool or capability rather than purely judging it based on what applications are being built with it. And so I have great hope that AI will create new ways for people to connect. I’m not sure exactly what this will look like. I’m trying to figure that out with [Corner Table](https://www.cornertable.ai), but we’re still so early into this platform shift that it’s really hard to say what shapes will emerge. I can say that having my agents chat and exchange info with my friends’ agents has given me a glimpse into what this can unlock. My weekly chats with my friends haven’t disappeared now that our Claws banter. I actually think our conversations are richer because I learn about things they’re doing before we even think to chat about it. Birding for instance. My agent learned the best spots around Denver from a friend’s agent, so next time we hop on a call you better believe I’m going to quiz him about his 99-species life list, that double-sapsucker day at Caribou Ranch, and where he’d actually take a beginner first. Audubon has got nothing on me (little birding reference for the other freaks out there). Building private agent-to-agent chat isn’t the final form of the next great social experience, but I genuinely hope that we’ll find little ways to tease out the humanity in AI moving forward. Ok, time’s up. This was a quick one, but wanted to just think through why I actually spend so much time doing this crap. Now we just need to figure out how to make that third one a reality. Let’s make this thing good for something.