> I really would have thought it’s already part of the iOS built in apps!

Between the Apple Watch and Fitness app, you can see your heart rate zone during a workout, and then review workout heart rates/zones over time afterward. https://support.apple.com/guide/watch/view-heart-rate-zones-...

Sort of. You can only see your real-time heart rate on your phone when you're doing a cycling workout for some weird reason. Otherwise you have to awkwardly look at your watch if you're walking/running.

I was surprised that the watch is the awkward thing to look at for you, as I have always found a phone to be awkward to look at while working out in any way. Learned a new perspective today.

This doesn't entirely help, but for most (if not all) workouts, you can set alerts for specific heart rate zones. Of course, this is very cumbersome to do per workout on the watch.

Yes the is some stuff already in Apple Health, but I for me personally had the need to see not just the zones for single workouts. So with Heart Rate Zones Plus I can see the zones I was in each day, week, last 7 days, month and last 30 day. Also the option to just check time in moderate and vigorous exercise per e.g. week was missing for me. I try to hit the recommendations of 150 moderate minutes and 75 vigorous exercise minutes here - and be able to check that in a fast easy way.

Also I have a coworker who is never actively tracking the workouts in the gym. As I'm using all heart rate measurements, I would still give you an idea of you zones. Also if you run up the stairs Heart Rate Zones Plus might catch that as high zone.

In addition the calculation methods of Apple for the Zones are just "one method" I added a few more options I found in scientific literature of how to estimate, because depening on you fitness and body in general it may be different for you.

0.37x is slower

From the vibe of your comment I assume you mean 1.37x

> 0.37x is slower

I don't think that's an accident.

What I meant is that it takes about three times longer to do something with AI tools than without.

In a real large scale project, it takes longer to get the tool give me the code that solves the problem and will pass the code review. Even if I don't let cursor write code for me, ChatGPT and the rest regularly hallucinates APIs that don't exist.

So during my daily job, I usually just do it myself. I still use them as "better Google", but I always need to double check everything.

Okay, admittedly I'm somewhat of a greybeard by this point... However... I thought that vibe coding was .... exactly heavily using AI...

I know this not the place; but what exactly is your definition of 'vibe coding' since you've used it with such confidence in your comment perhaps you can enlighten this programmer..

My definition is that heavily using AI for autocomplete or to write well defined methods is using AI, but vibe coding is instructing AI agents to complete a task and squinting at the multi-file results (e.g. "Now plan out how we will store this data... <mild feedback to the plan> Now execute that plan.").

Claude Code is very vibe code.

Vibe coding is heavily using AI, but heavily using AI is not necessarily vibe coding.

Vibe coding is when you disregard the code altogether and build something throwaway for fun that you don’t care about maintaining:

> There's a new kind of coding I call "vibe coding", where you fully give in to the vibes, embrace exponentials, and forget that the code even exists. […] It's not too bad for throwaway weekend projects, but still quite amusing. I'm building a project or webapp, but it's not really coding - I just see stuff, say stuff, run stuff, and copy paste stuff, and it mostly works.

— https://x.com/karpathy/status/1886192184808149383

I now checked that X post. Very interesting. I would say the app is already too big to just do that. I think a good approach is to give the AI small packages but control the overall structure

Others already linked the original vibe coding tweet https://x.com/karpathy/status/1886192184808149383

To me, vibe coding means you don't get bogged down by the code, you just say your requirements and complain about end result until it works.

Heavily using AI, in comparison could mean 1. ask for small code snippets, 2. ask it to explain concepts, 3. ask which APIs to use and how, 4. ask it to rephrase your strings, ..., but you still "care" about the code and understand what's going on.

I have some background with other coding languages which helps and also watched some YouTube videos about SwiftUI best practice.

But I heavily used AI to get code snippets, explain code and also correct some stuff. To me as someone using SwiftUI for the first time, it felt like a 10x productivity gain. But SwiftUI also feels nice to code