How do you know when to call it quits?

Daniel spent his day studying TensorFlow at the library, then prototyped a chatbot for workout plans as a step toward his "Airbnb of gyms" idea. He discusses Seth Godin's "The Dip" and the sunk-cost fallacy, then shares how he quit working at 5:30 PM when fatigue hit, framing it as courage rather than failure.
Chapters
What's going on, girls and boys? Welcome back to day 91 of the Daniel Bourke Show. Thank you so much for tuning in on the podcast, on the YouTube. I hope whatever you're up to is an awesome day. I've had an amazing day. I'm pretty wrecked at the moment, as you can probably tell by my face. I'm probably extremely— I've probably got bags under my eyes. I've been looking at a computer screen for far too long, and once again I've made the mistake of leaving this podcast to the last thing I do in the day, which I've learned in the past Daniel, stop doing that. Stop doing that. You should get the podcast done early in the day so you've got more energy and you've— you're not so tired and you don't have a belly full of food and you're not about to have a shower and go to bed. But nonetheless, we're here once again. I've taken 91 days of this, this lesson and I still haven't learned it properly, but that's alright. This is the last thing I am doing for today. What did I do this morning? I got up, I Around 8 o'clock, I'm adjusting my body clock back to waking up at about 6 or 8, waking up with the sun. And then what did I do? I went to the library, actually. There's a new library. I dropped my little brother at school, and there's a new library that's been built not too far from my house. Well, refurbished, actually. And I got about 4 or 5 hours of studying there, maybe less, maybe more, I'm not sure. But I like studying at the library. I haven't done it in a long, long time. I only did it when I was in university, but I think it's definitely the environment. Like, I felt like I was more concentrated just because I was in the right environment to get study done. Recently, most of the time I'd be doing my work at home, but sometimes that's distracting because I've got my family walking around asking me questions, which is not a bad thing, not a bad thing, but if If you're, I don't know, in the zone and you don't want to be disturbed, sometimes I just put my noise-cancelling headphones on and just— I'm rude sometimes. I don't answer people because I'm just in the zone, like in the study zone. And— sorry, someone just rode past and distracted me. They're probably wondering, what the hell is this idiot doing in their car with the lights on with a camera in front of his face? Shh, he doesn't have to know. He can come on the show if he wants. But yeah, I think the environment really, really set me up for success there, and I got some— I got some— a lot of work done. I got a whole class done in my deep learning course that I'm doing, which was pretty intensive. It— we talked about— I learned about TensorFlow, which is— TensorFlow is a library for for deep learning that's developed by Google, and it's very comprehensive, and it does a lot of the, the groundwork for you in terms of forward propagation, back propagation, error calculation, etc., etc., which is a whole bunch of jargon, I know, but essentially it helps you create machine learning type software. Let's just put it that way. And I'm probably butchering that, so don't take my word for it. Look up TensorFlow if you want to find out more about it. And then what else did I do? I looked into making a chatbot. And what's a chatbot? So if you've ever used Facebook Messenger, you can discover bots on there that will chat automatically and respond automatically to different things. So I'm thinking about making a chatbot for workouts that'll help people learn how to train and like give them different options of exercises for different body parts and different workout plans and whatnot. So I'm thinking about making one of those. Well, actually, I am going to make one of those just as a side project, and then eventually that will build into the overall arching goal I'm trying to, trying to make is a— the Airbnb of gyms. Let's just imagine that. So if you go on here and you can find any fitness facility around you and use that fitness facility. And so I'm gonna start it off, I think my intuition is telling me start small and then build up bigger and bigger and add features. So I'm gonna start with a Messenger bot that helps people work out and then build it up from there. So that's just my intuition and that's, I don't know, I've got my eyes on the overarching goal, the moonshot is to make the biggest fitness facility in the world by connecting everyone connecting all the fitness facilities on the internet. And I know psychologically this is doing damage to me just about talk— just talking about it, because it's one of those cognitive biases, um, that I've spoken about in a previous episode, that if you talk about your goals, uh, you actually— psychologically you think you've already completed them. So I like to keep them hush-hush, but I also like to publicise them so people go, "Oh, why aren't you working on that? Why aren't you working on that?" And I think I've faced some resistance in the past. I think that's what yesterday's episode was about over this, this massive goal. I've been taking on so many different things. And today I was studying for about 4 or 5 hours and then finished up. My brain was fuzzy by the end of it. That's when I can tell I need to stop, when my brain gets fuzzy and I can't comprehend anything. I need a break. I need to go eat something or have a nap or something like that. Let it all sink in. And even though I was in the fuzzy mode, I started to build this, this messenger bot using chatfuel.com. You should check it out if you want to make your own bot. And it was like a, well, like, what am I trying to say? What word? What word? I don't know. A wealth of energy, let's just say that, came down upon me because I was doing something that I was fundamentally interested in. And I feel like that's, that was that secret sauce, you know, that sauce that gets you out of the bed in the morning. Like when you have an event that you're really excited for, you just jump out of bed, whereas on other days you sort of mosey on out of bed because what you're going to do that day is not as fun as staying in bed. It's not as comfortable as staying in bed. And I think it's good to have a bit of both, actually. You've got to have that thing that gets you out of bed, and you've got to have those days where you need to have that self-discipline to keep pursuing with your goals, because it's not always going to be up. You're going to have those up moments. Then I think Seth Godin has a book on this called The Dip. You're going to be up, up, up, up, up, up, dip, and then back up, up, up, up, up, up, up. Um, And I think I've done a few of those cycles in my lifetime. I don't know where I'm at at the moment. Maybe in a little bit of a dip at the moment, I think. But nonetheless, I'm going to keep pushing through, keep marching forward. And like today, for example, I reached a dip at about, I don't know, 5:30 PM. My brain was fuzzy. I had on my to-do list for today to write an article about Part 2 of the How I'm Learning Deep Learning series, and I got, I don't know, probably 30% of the way through it. I've done the outline of it, but I just, I spent about an hour on it and I couldn't, I couldn't produce any more. Like, my brain wasn't working. I need a break, I need a break. So I went and worked out, I came back, trained my little brother because I'm training him up to be a A rig is going to be shredded in a few months' time, if I have my way, that is. But I called it. I called it for the night, I think. I'm not going to go back and finish it tonight, even though it was on my list for today. Sometimes courage is— I think Tim Ferriss says this, or someone else says this— sometimes courage is calling it quits. So I know I'm not going to put out some great work tonight if I keep going. Instead, I'm going to go back, read some more of this book that I've been reading, The War on Art— War of Art, sorry— and pass out, go to sleep, wake up early in the morning, and I'll get cracking on it first thing tomorrow. So tomorrow you'll see an article from me, How I'm Learning Deep Learning in 2017 Part 2. And yeah, then for the rest of the day, I'll either be teaching my little brother or doing some study. The study is probably definitely in there somewhere, but I think, yeah, that's the daily challenge. What could be the daily challenge? I think that's a good wrap-up. Sometimes you got to call it quits. You got to know when to stop. I'm definitely one of those people to set incredibly ambitious goals for myself for one single day. I think people overestimate, me included, what they can do in a day. But underestimate what they can do in, say, a year or something like that. And I, I'm definitely, definitely one of those people. If you— if I look back at the years that I've gone through and you'd have told me at the start of the year I would have done this by the end of the year if I kept going, I wouldn't have believed you. Same thing, um, with a day. I always overestimate how much I can get done in a day. So I think I may need to just a general rule for me, maybe 2 times the amount of things I can get done, I can think I can get done a year, and maybe half the amount of things I can get done in a day, and see what happens. And I know I'm a contradict, like I'll say, oh, you got to keep working, you got to keep pushing, you got to keep pushing, why can't you do more? And at the same time you're like, oh, maybe you should limit how much you do in a day. I think you got to pull from both ends. You just got to have those days, one where you just go balls to the wall and you you've completed like, I don't know, 200% of what you wanted to do, and then other times you'll realize, okay, I'm only going to get 50% done of what I wanted to do, but I've done that 50% 100%. So I think that's, that's more important than worrying about, I don't know, doing everything that you said you'd set out to do. So, spend time effectively. And efficiently. Who knows? But yeah, knowing when to call it quits is a big thing, as well as that's another cognitive bias, the sunken cost fallacy. It's where you get too deep in something that you're doing and you just keep going with it anyway, even though continuing on with it is detrimental. Is not even— you're not even getting any benefit out of it, just the fact that you keep going. That's a sunken cost fallacy. That's another cognitive bias. So what's the daily challenge? I would say reflect on something you've been doing lately that you may have been overestimating your ability, or if you've got a long-term goal, are you underestimating your ability? So if you've got a 5-year plan, why can't you do it in 1? Or if you've got a whole list of things you want to get done for the next 3 days, why, why do they have to be done in the next 3 days? Could you cut that list in half and just do the 50% of things that will matter the most? So I would say, yeah, you spend 10 minutes reflecting on where you're overestimating what you can do and where you're underestimating what you can do. And yeah. That's the daily challenge, but I'm just going to wrap it up. I'm going to go upload this podcast, have a shower, get into bed, start reading this book, and then go off to dreamland. We'll see you back for day 92. Catch you later.
More from Daniel

Small Language Models (SLMs) Are the Future: Fine-Tuning AI That Runs on Your iPhone

Introducing Sunny - A MedGemma-Powered Skin Health Tracking App (Kaggle Competition Entry)

Benchmarking: NVIDIA DGX Spark & NVIDIA RTX 4090 | Part 4 - Comparing the results

Benchmarking: NVIDIA DGX Spark & NVIDIA RTX 4090 | Part 3 - Running LLMs on the RTX 4090

Benchmarking: NVIDIA DGX Spark & NVIDIA RTX 4090 | Part 2 - Formalizing the tests

Benchmarking: NVIDIA DGX Spark & NVIDIA RTX 4090 | Part 1 - Creating a series of tests

Local Multimodal RAG on the NVIDIA DGX Spark | Part 4 - Publishing our demo to Hugging Face

Local Multimodal RAG on the NVIDIA DGX Spark | Part 3 - Putting it all together


