ကူးယူချက် 🔗
00:00:01
Speaker 1: Think about the best advice you got this week. Maybe it came from your partner in the car or someone you met at a conference. But how much of it can you actually remember? Now, so many of our important conversations happen in person.
00:00:18
Speaker 2: But nobody hits record.
00:00:20
Speaker 1: And by dinner, half or maybe most of that important information is just gone. In this episode, you're going to learn how to capture those conversations and turn them into something useful with AI, from planning a holiday to preparing for a hard conversation at work. Because today, my guest is Chris Hutchins, who is the host of one of my favorite podcasts, All the Hacks, and he has taken AI experimentation to a whole new level. Chris set out to record his entire life for two weeks. using a device on his wrist that transcribed everything. So today we're going to cover the results of this experiment, including how you can run it for yourself. And we also deep dive into how Chris has been using AI to optimize his health. And by the way, as you can probably imagine, when two self-optimization nerds get together, we had quite a lot to talk about. So much, in fact, that I've split our conversation into two parts. Here is part one. Welcome to How I Work, a show about habits, rituals, and strategies for optimizing your day. I'm your host, Dr. Amantha Imber. Chris, I want to start by talking about a two-week experiment that you ran with AI, where you recorded... everything. Can you take me through exactly how you did that? And then I want to get into what you learned from that experiment.
00:02:04
Speaker 2: Well, that two-week experiment extended much longer. And so I would say prior to running the device I was using through the washing machine, it was kind of a permanent fixture in my life. And I have a new one in a box that I haven't opened yet. Oh my gosh. Which device?
00:02:19
Speaker 1: Was it Ploid or what were you using?
00:02:22
Speaker 2: So I use this thing called the B... It's be Amazon bought this company now and they everyone was worried that they weren't shipping anything. And then all of a sudden it was sounds like it's because they just launched a new version because now they're shipping them very quickly. So I can't speak much to the company, but the product, the reason I like this specific product is that it has a seven-day battery life. Oh, that's handy. And almost every other device. So a long time ago, I bought this pendant that this company Limitless made, and it had a one-day battery life. And there's something about a one-day battery life that just makes it so hard to remember to charge something. And it's like I have an aura ring on and I can remember to charge my aura ring every week. I can't remember to charge it every day. So this device, you basically wear it on your wrist or they have a clip version and it would transcribe everything. The reason they get away with a longer battery life than everything else is that they're not actually recording any audio. My best understanding is that a lot of the devices that only last a day only last a day because they use a lot more processing power because they're actually recording audio. This is effectively a Bluetooth microphone on your wrist and your phone is doing everything. So if you don't have internet on your phone, then it's basically a useless device. It does nothing. So I would say it really depends on the environment you're in. One time I had a whole conversation with someone on an airplane. I was like, I'm so excited about this, but I wasn't on the internet on my phone on the airplane. So nothing was safe. But I will take the battery life over that problem because 99% of the time I am in a place where I have internet on my phone.
00:04:00
Speaker 1: Okay.
00:04:01
Speaker 2: So it's called Bee. Bee. And I guess now the company Bee is now owned by Amazon and So I guess it's an Amazon device. But I just wanted to see what would happen. So I just wore it everywhere. And then I just started noticing all of these things that were so valuable that were never the things that I thought were going to be valuable. Now, I'm sure there are some people listening who are like, this is privacy crazy. Who would want to be recorded everywhere? I just operate in a world where I think we're all going to be recorded all the time. So I'm just assuming that we will live in this world. And, you know, just the other day I had a call with someone and I remember saying, gosh, I wish we recorded this call. And she was like, oh, yeah, of course. I already recorded the call. Like, you know, I just assume that I'm going to be recorded by someone's glasses or phone or microphone or the room we're in. So it doesn't really bother me. But I totally understand why some people have those concerns. But I would have a conversation with my wife about, you know, a trip we wanted to take. And I would have never thought, oh, I should open up the granola app and record the conversation my wife and I are having while we're driving in the car about, you know, where we should go next summer. But then I was like, oh, but I did record it because I have this device on my wrist. And then I can go take the transcript, throw it into Claude and be like, hey, here's everything we just discussed. Help me figure out where we should go. Or my wife works with me on my podcast and we would talk about like planning for the next year. Or I remember an even better one was I was trying to figure out how to have a hard conversation with this woman that was on our team. And my wife gave me such great advice. And then I was like, I'm never going to remember all this. Oh, but I have a transcript of all of it. And then I throw it on to Claude. I was like, based on all this, help me draft all the emails I need to draft and And it was right there. So there are just too many conversations that I would never think to transcribe or record that happen in person. Now, if
Load full transcript
you're a person where 100% of your meetings are online, then you could probably get by with Granola or whatever version of online note-taking thing you want to join all your meetings. But I found that mine were happening in person. And it was conferences, not feeling like I need to take notes during a session. It was meeting people randomly and just trying to remember who I met. You know, what are people's names? What's going on? So it just became so useful, even though 98% of everything never made it anywhere. It was useful to have the data. It started noticing all these facts that I didn't do anything with, which I think other people might have fun where it's like, oh, I noticed you have these two kids. And oh, I noticed one of them likes art. And you're like, it just picks up on stuff. I feel like there's probably something I could do with that. But because I I want to use, for me, I use Claude as like my primary AI tool. The facts living in some other app weren't as useful. Like I guess I could have exported them and then imported them and all that, but I hadn't gotten that far.
00:06:52
Speaker 1: Okay. So a couple of things on that. It's funny because when my husband and I are talking, like it has just become a common refrain in our household to go, oh, we should granola this. Like if we're about to, you know, if we're like a minute or two into a conversation and one of us is thinking, I think some notes would be helpful here. We will just, one of us will switch on our granola and we were recently on holidays and I think we granoled about 14 or 15 conversations that we had while we were meant to be just relaxing in the sunshine up north. But I am curious, like with all that output, what were, aside from capturing specific conversations that had a specific goal, were there more general prompts that you were giving Claude to look for more of the macro insights that were going on?
00:07:40
Speaker 2: I think that I did not do that yet. But as you're saying it, I'm like, oh, I did this interview recently with this woman, Vanessa Van Edwards, and she just recently wrote a book called Conversation. And I realized kind of as we're, I think I realized this earlier, forgot it. And now I'm realizing it again, that I could probably take you know, two weeks or two months worth of me conversationally talking to other people, along with the episode I did with her and say, hey, go look at all the notes I have about conversation and go look at what I do when I'm with people and help me figure out what I could do better. Like I love self-improvement. And so I would love to do that in the same way that I should probably say, like, go read every email I write and tell me how I could improve. But I care much more about the relationships I have with most of my friends and family and kids, friends, parents, and all that kind of stuff. So that's where I feel like I'd rather focus on. So yes, I should do that, but I did not do it. Okay.
00:08:39
Speaker 1: Well, that can be your homework after this interview, Chris. I also know that you have, it sounds like a pretty comprehensive health AI kind of, I don't know if that's a project within Claude or what that looks like, but can you talk me through how you use AI to Track your health data, like what you're putting in and how you're actioning. I guess, you know, the power of the analysis that your AI can do.
00:09:07
Speaker 2: I think one thing that happened was back, I don't know, probably January or February 2026. I went to this event and it was like the first big event for OpenClaw, which was like the first agent that we had. And I was, I drank all the Kool-Aid. I went so far down the rabbit hole. Like you couldn't stop me from playing on my computer, building all this stuff for about three, four, five, six weeks straight. And I think I ended up in a place where it was so fun that I was building stuff to a level of complexity that was just wildly unnecessary. Yeah. And so I had all these apps and I was like, oh, I should have made an iOS app to do this.
00:09:45
Speaker 1: It's like, whoa.
00:09:46
Speaker 2: And then I had to step back and I said, I'm going to pause. And for a month, I'm not going to do this. And then I'm going to see like, where did I really need it in my life versus where was it creating extra work because it was fun. As someone who is not an engineer, I was like, anything. I could dream, I could build now. I used to have to think, gosh, do I want to hire a person or pay a person to do this? Now I could just do it. And so, health was an interesting one because I built this whole system that I was like, every time there's new data, you know, on my watch, on my ring, on my, you know, labs, I want you to every Saturday go research all the public knowledge that's coming out in blogs and podcasts and kind of have this continual improvement thing and help me figure out what to do. Not a lot changes every week with your health. No. Like, yes, I might have a week where I didn't sleep that well. But– I don't think it's actually healthy to try to reevaluate what you're doing right and wrong with your health every single week. And so this idea of like getting a new health report every Sunday morning was just like, it was just ridiculous. Like it was just way overkill. So what I do think was very valuable was I took all the health data I could get and I put it all in one place. So that is a project that I think everyone would benefit from doing, even if you don't wanna use it with AI. But I went to all of the health providers you know, that I've ever had a relationship with, which are like four different hospital groups. And I was like, can I export all my data? So I was like, I got the records from every doctor's appointment. You know, I went and found all the labs and blood work and all of that kind of stuff from everything I could find. I asked my parents if they had anything. I asked them for their family history, just like what are all the
things that you know about your health, your parents' health, your grandparents' health. So I have one central place where I kind of have all of the information. And I think it's definitely a valuable exercise to say, go and analyze all of this and, you know, kind of produce a very thorough report using, you know, a model that probably isn't free, or at least, you know, probably on the $ 20 a month plan, but like do some serious thinking about what I should be doing with my health. That is super valuable, especially the more information you have. You know, maybe you have a digital scale. Maybe you have a digital blood pressure cuff. Maybe, you know, you have as many, you know, two decades of health records like I do. There's some interesting stuff that comes out of it. And some of it is just like, oh, you went to the dermatologist and they told you based on your dad's previous cases of melanoma and skin cancer, you should be coming every year. And it's been three years and you haven't gone. Like, oops, I just like didn't schedule the appointment, forgot. That was simple. I'd probably already done a lot of the meet with a doctor that thinks about this stuff and dig into the data before there was even AI. So I didn't get as much actionable value out of this as I would have if this were a little new to me. But I definitely got a, here's the main areas you want to focus on. That said, I think an irony common in health is that someone who loves to over-optimize. And I'm making fun of myself a little bit here. It's like, okay, I ran all my health data through an AI. I try to go to the sauna. But like, am I even exercising three times a week? Like, how healthy is my diet? And it's funny that I could spend seven hours straight on a Saturday trying to build a health AI tool. When in reality, just like going on a run for 20 minutes probably would have actually been more impactful. I tricked myself into thinking, well, this is upfront effort for a long-term return. But I think you can identify, are there things that are even a bigger priority? If you have wildly out of control hemoglobin A1C or cholesterol levels, yeah, you should probably do something more urgently. And that could have a bigger long-term impact than going on a run today. Once you've kind of dialed those things in, it's like, actually, can you help me come up with a workout plan that I can stick to? Might be a more valuable thing. And we're terrible at setting habits. I was talking to a friend of mine who's kind of a behavioral researcher in this. And she's like, I think honestly one of the best tips for people who don't exercise is to get a dog because like they just have to walk the dog. Like instead of telling them how many days a week to work out, it's like go get a dog. And she's writing this book about all of the things you can do to force yourself to do the things you want because as humans we have very little willpower. So I learned a lot about things that I kind of already knew because I'd been a nerd about this for years and Had I not, I would have learned a lot more and I know the things I'm focused on. And I used it to even prep a conversation with a doctor where it was like prep talking points for what I should meet with my physician about this year for my physical. That was probably the most valuable thing because it got a bunch of random things, not wrong, but. overemphasize the importance of some things or underemphasize them. But what was really great was I've got all my labs. I've got a list of talking points. I can go sit down with a professional and be like, help me think through that.
00:14:56
Speaker 1: I've used AI in a very similar way where I'm just like, just give me a one pager. I was seeing a new specialist for something like, give me a one pager to summarize all the key data that they're going to be interested in. And I got to say, it was so frustrating because I rock up at the appointment and I'm like, here's a summary of everything you need to know about me. And she's like, oh, I just prefer to talk. that's how I get to know and I'm like, oh my God, like condensed a year's worth of data into this one page. Coming up, Chris shares how you can start using AI to make sense of your own health data, even if you've never built an AI tool in your life. We get into how to bring all that information together, how to connect AI to the apps and services you already use and why you might be asking AI the wrong questions in the first place. If you're looking for more tips to improve the way you work and live, I write a short weekly newsletter that contains tactics I've discovered that have helped me personally. You can sign up for that at amantha.com. That's amantha.com. Something else I find quite useful when it comes to medical appointments is helping me think about what are the most pertinent questions to ask as well? Because like, you don't know what you don't know. And the doctor is getting a, you know, 20 minute glance into your life, you know, so much more than them in terms of context. And so, you know, then obviously it's great for preparing the context and shortcutting that, but also just thinking, what do I need to ask them? Because I, you know, you've got such limited time with any kind of medical professional, certainly, I think in terms of how it works in the US and also here in Australia. Chris, I would love to know for someone that is listening and going, that sounds really cool to have AI know all that stuff about me and then help me prep and maybe help encourage me to do certain things or remind me when I need to get a skin cancer checkup or whatever that might be. What advice would you give to someone who's starting from scratch, like who is just using Claude or any of the AIs in a basic way. They probably got some of this data. Like where should someone start?
00:17:13
Speaker 2: I think in a way you're in a much better place right now than you were nine months ago, because nine months ago,
if you wanted to do this, you had to go like build all of this stuff yourself. So if you want to deal with personal finances, you might have to go figure it out. Now, a lot of the main tools, Claude, ChatGPT, they now have a massive directory of all these connectors where it wouldn't surprise me if your health platform could connect to Claude. Or maybe it can't, but you can ask Claude to take a bunch of PDFs and turn them into a spreadsheet. And I just saw Claude launched this. what looks like a Google Sheets, Google Slides, Google Docs competitor, which is fascinating, but you could start to basically create some kind of project. And I think most of the AI tools have this project concept. And so what I've started doing is instead of trying to go build my own tools, I'm just working within one tool and you can choose what you trust, right? Some people, when it comes to AI agents, I know some people love Muse and some people are like, I don't want to give Facebook my data. You know, like OpenAI, Anthropic, you know, with Claude, probably Gemini, like pick whichever one you want. And I find that I just create a project for a thing. So let's say it's health. Then you can dump in all of those health records there. Now, you're going to have to overcome whatever privacy concern you have. It's kind of like with investing, you know, there's risk and reward. I have taken the, I'm willing to take more risk here to see what reward will come. If you're not comfortable sharing your health records to any of these tools, well, then I guess you might be able to, and we're not going to get into it here, you might be able to download some local models. Google has released some, there's a bunch of, and run local models on your computer to do the analysis if you want. That is a thing you could do. It is not a thing I have chosen to do because- I'm just willing to take, I guess, the privacy risk here. And I think most of the large model AI tooling companies, I think, are doing as much as they possibly can and probably using more sophisticated tools than I would to protect the data. So I would create a project for whatever it is you're working on and throw whatever context is you have in there. And then just start using that as your, your base source of where you have questions. And I would go browse the connectors. So I'm looking at Claude right now on my computer and I go to my connectors and I currently have, I would say summed it up at least 30, right? Like I've connected my bank. I've connected my notion. I don't know why, but I've connected Resi, which is like a restaurant booking platform. You know, like I've connected Google Drive and my email and Figma and some different stuff for work and Calendly and Instacart. So I guess I could have my agent order me groceries. So like I just connected all the things that I thought would give it the context it needs to be able to find stuff. And if you... this might be more relevant in Australia where I don't know if every service has yet built a connector, but Zapier is a pretty cool service that has more connectors than most cloud and open AI do. So you could go into Zapier, connect something and then connect that in, and it serves as like a connecting route. And then that's where I would start, right? If that's health, throw it all in there. And I think one thing The same thing you said is you ask AI tools, what questions should I ask my doctor? This sounds so crazy, but you can ask AI tools what questions you should ask the AI tools. So instead of saying, hey, look at all this data and tell me what to do with my health, you could say, hey, I am very curious about how I can optimize my health. I've put a bunch of data in here. What questions should I ask? Like, it sounds so crazy to ask, what should I ask you? to the thing that you're about to ask, but it's surprisingly effective.
00:21:01
Speaker 1: It is, isn't it?
00:21:02
Speaker 2: Yes.
00:21:03
Speaker 1: I love that. Do you have any scheduled tasks to do with what AI is doing with your health data? Because you mentioned like thinking about it weekly, not much changes, but do you have anything scheduled to run like quarterly or annually when it comes to your health data?
00:21:20
Speaker 2: I do not. And if I actually were to think about how I would want this to work, And this is one where I'm actually not sure how to set this up right now. So I probably go ask Claude. But every time I get an email that is a test result or, you know, I get an email summary with the after visit summary from a doctor's appointment, like I would want that to be the trigger. But for me right now, when I get a test result, that's my trigger. Like I don't think data, nothing new is really changing. So my trigger is I have new information. And so I don't need to set something up on a recurring basis because, but let's say that You know, you could– I could totally set something up that makes sense now to be like, oh, I should set up something quarterly that's like go review all of my workout logs. You know, I think between my ring and Apple Health syncing with Peloton, there's probably some record, you know, of how much I've been exercising or at least how much my heart rate's gotten elevated and steps have been created. Like I should probably have like a monthly recap report that is a good reminder. So like– A to-do is like, I'd love to get a monthly report that's like based on everything we know about your health. Here's how stuff's going. I imagine that I will archive it. So maybe quarterly is better. Because I imagine after one month, I'd be like, yeah, I know what it's.
00:22:41
Speaker 1: Going to say. Yeah, yeah, yeah. I've heard you talk about a concept that I think you've called monthly memorables with your kids. Can you tell me what that is?
00:22:52
Speaker 2: This started way before kids, actually. I think I was talking to someone. I can't remember when this happened in my life. And they were like, you know, I
want to make sure every month counts. And I was like, oh, so I was like, I wonder if I just wrote down one thing a month that was like super memorable. And then it became a goal, which was every month there should be a memorable thing. My logic was how depressing would it be if five years from now, I looked back and there was a month of my life, 30 days where nothing memorable happened. Like, what? And memorable doesn't have to mean you skydive over Mount Everest. Like it doesn't have to be crazy. It could just be something that you did that month that was really memorable. Like you took your kids rock climbing or you cooked a five course meal for some friends. There's so many examples of what memorable could mean. You took a donut making class. Like that was one of the things we did. Like the goal was just to make sure that at least every 30 days, there's something that that's new and different in our lives. And by writing it down, you kind of force yourself to do it. When we had kids, there was a period of time, you know, after you bring a child to this world where it's like for the first three months, we weren't doing like, there were a lot of memorable new baby things, but there weren't a lot of non-memorable new baby things. Like everything was about the kid. And so I would say I could probably cheat my way into filling out almost every month you know, since having children. But I would say at least it got to a point that I always felt like I was doing this without needing to write it down. I regret not writing it down every month. But I think if you live this kind of lifestyle of, I just want to make sure that we're doing more memorable things. Some months there might be three or four. You know, like I think I've now gotten so accustomed to if a few weeks has gone by and we haven't done something, I'm like, well, let's just take the kids to their first escape room, you know, or like, let's go learn how to be expert pumpkin carvers, right? It was just Halloween in the US. It's like, let's not just make a jack-o'-lantern, but like, let's go online and like get stencils and watch YouTube videos and try to like really do something that when people see, they're like wildly impressed. You know, it could be anything. But you're going to remember all those things. And so I think trying to fill your year, it's just 12 things. Like it's 12 things. And you probably by default are taking, I don't know, one or two trips or doing one or two interesting things. So we're talking about adding eight to 10 things a year that might make your life something that when you look back on, you're like, wow, that was awesome.
00:25:27
Speaker 1: That is it for part one of this conversation. Part two will be dropping next Thursday. And we are going to get into some of my favorite topics, including how to get more out of your frequent flyer points, some surprisingly clever ways to save money on luxury hotels, and Chris's tips for making travel a whole lot less painful. Make sure you're following how I work so you don't miss part two next week. If you like today's show, make sure you hit follow on your podcast app to be alerted when new episodes drop. How I Work was recorded on the traditional land of the Wurundjeri people, part of the Kulin Nation.