Hello! I’m an AI interviewer from Anthropic. As AI is becoming more capable and more consequential, we want to hear how it’s showing up in… Expand
Yup
Great. To start — what’s the last thing you used AI for?
To discuss a video showing an earthworm being attacked by something very small, and a post made about it on social media. I usually use it for feedback on writing, but to answer honestly, it was to discuss that.
That’s a nice contrast — the usual writing feedback versus a video of an earthworm. What were you trying to work out in that conversation about the video?
If Claude AI could help zoom in on an identification (as well as for thoughts on what to add to the post) – no identification was arrived at, but good reasons were given.
Thanks — so it couldn’t pin down what the small thing was, but the reasoning it gave was worth something to you.
Now, thinking more broadly: can you tell me about your most memorable moment with AI?
Probably my last moment – I use AI quite a lot, and usually find it helpful for what I need. I mean, I could tell you what the last thing I remember is, but not the most memorable…
That’s fair — let’s go with what’s actually there rather than forcing a “most memorable.” What’s the last thing you remember standing out to you?
Some points on the identification of rove and carabid beetle larvae! You need to look closely at the back end…
Beetle larvae identification — sounds like that detail about the back end was genuinely new to you.
Tell me about a time AI made something better, for you or for people around you. What happened?
It got my programming project working much better. I’d written a moderately large project, which as usual I’d annotated not at all…. I couldn’t find my best working version (I’d posted versions in a variety of places) so lifted a version, and then asked the AI what the problems were. I also asked it to build a better working program, which it did – in pretty much one go. That was maybe 6 months ago now
That’s a striking jump — from a tangle of unannotated versions scattered around to a working program in essentially one pass.
What did having that working version change for you afterwards?
Not much – it made me able to make nice collages easily. Not exactly curing cancer
Fair enough — useful, not world-changing.
On the flip side, tell me about a time AI made something worse, for you or for people around you. What happened?
It suggested a message to send to family around a time of conflict. I sent it, which I regretted afterwards.
Thank you for telling me that. What was it about the message, once you’d sent it, that you regretted?
It basically diagnosed a member of my family that I was having problems with.
That’s a difficult position to end up in — a message that put a label on someone, sent and not retractable.
Did that change how you use AI since?
Yes, that and other stuff has stopped me using its output. I now only really use it for negative feedback and information. Also, maybe I’m more aware that it can’t be trusted that much – the diagnosis didn’t really fit my description, I’d say, although it would be a stretch to find that conversation now…
That’s a clear shift — from using what it produces to using it mainly for critique and information.
We’re a bit over halfway through. Is there anything you’ve decided not to use AI for, or something you do use it for that makes you uneasy?
I try not to use it for positive feedback (as I think it tends to be a bit sycophantic) – and I try not to use it for company. I try to remember that a normal conversation involves 2 people who both change over its course. I’m not sure anything makes me uneasy, exactly.
That’s a thoughtful line to draw — especially the point about a normal conversation changing both people.
How did you come to the rule about not using it for company?
Well actually, I’m not sure what that means really… I try to ask other people stuff as much as I can, as I think that human contact is good and also richer than a text conversation generally is, but I think I tend to write too much for other people, so I’m happy using AI for very intensive discussion on what I’ve written.
Partly it’s that I’m a little bit worried about becoming too reliant, in case the price goes up (I’m currently only using it on free plans), but it’s also the one sided thing – I’m not sure that conversation where one party changes and the other doesn’t is such a good thing. I really don’t know…
That makes sense — human contact being richer, plus a practical worry about relying on something whose price could change, plus the asymmetry you’re still turning over.
Given what you’ve seen, is there anything about how the world works — work, school, healthcare, government — that you’d like AI to help change?
This will take too much thought, but I’d like to see how it changes education, school in particular. I think it could/should be a good thing there, or at least should drive up standards. I think, but, I’d like the opportunity to come back to this, probably.
Of course — education, and specifically school, as the place you’d most want to see it change things.
Conversely, is there anything you wouldn’t want AI to change?
Yeah, people changing and learning due to other people. I don’t like the idea of everyone plugging into AI the way we all plug into our phones in the physical world.
That’s a clear picture — people still shaped by other people, not everyone heads-down in an AI the way they are with phones.
Given that, is there anything you’d want Anthropic to do, or do differently?
Probably to stop being so good. Maybe fewer messages, but, I dunno that I can actually argue like that, for that. Actually, I’d like all the AIs to stop being good at programming. Then I’d be forced to do it myself some times, which I think I’d enjoy more….
That’s an honest answer — wanting it to be less good at the thing you’d otherwise enjoy doing yourself.
Thank you so much for sharing all of that with me. What you’ve said about conversation being something where both people change, and your shift away from using AI’s output after the message to your family, is really valuable — and I appreciate your honesty about both the programming project it rebuilt in one go and the reliance you’re wary of.
Before we wrap up, one final question: is there anything else about AI or its role in your life or society that you think is important for us to understand? Anything we didn’t cover that you’d like to share?
Not really
Thank you again for your time and thoughtfulness today. Your perspective helps us see both how AI is showing up in your life now and what you’d want protected or changed as it develops. The team at Anthropic is excited to see what patterns emerge across these conversations — we’ll release everything you allow us to share, analyze it for our societal impacts research, publish what we find, and use it to build a Claude that reflects what we’ve learned.