A back button for when you're in the exercise itself.

CONCLUSIONS:

1. Schedule function needs to recognize a baseline (forever) and a "this week's" schedule. Baseline is my "available" days and this week's schedule accounts for the immediacies and variances of life. I think this is 100% doable with a future block-change to the current app.
   1. Bonus points: when "this week" is entered (which by logical default is different than baseline, otherwise we wouldn't be entering it) Dr. M informs the athlete what the consequences are (i.e. how it will adjust the programing), impact to goals, and suggestions to mitigate the impacts.
2. Recognize that not all deviations to weight lifting programming are "bad" in terms of fitness. Spending 8 hours doing construction or hiking in the mountains is different than 8 hours attending an off-site conference. This might be reducible to just asking "since you're not sticking to our weights programming, will that day be physically demanding, and if so how much on a scale of 1 to 5?" If it's that simple, then it could be done within a block-change to the current app.
3. Stretch Goals:
   1. Move towards being a total-fitness app. I think current medical knowledge says just strength training doesn't sufficiently promote cardio health, and cardio health training without strength training is a one way ticket to decrepitude. I know that both my Whoop and my Functional Health doctor look at my data and say my Dr. M routines don't come close to stressing my heart enough. Dr. M. hardly ever gets me out of zone 1, and mostly in zone 0. And I'm hugely disappointed in Apple Fitness and Health; they're essentially libraries of videos, and data repositories without any checks promoting the data to "information", so they are not candidates to create a weekly mix of cardio and weights.
   2. Move from being an "app" to an "assistant" to being a "coach". That requires understand what each of those humans do and how the app would need to simulate a human to do it. It would also need to sense the user's emotional and intellectual levels so it could respond in a constructive way.

RATIONALE:

You and I've briefly talked about "wouldn't it be cool if all of these health and fitness apps, including Dr. M., were integrated enough to share appropriately." But that's not going to happen unless one of the big companies like Apple or Google sets a standard. So let's talk about how Dr. M. can be as supportive as possible by being as flexible and knowledgeable as possible.

Making broad generalizations, some people have fixed schedules. My life as an engineering undergraduate and graduate comes to mind: Every day within a semester has a fixed routine, so workout types are at fixed times on fixed days. Other people have a more chaotic weekly routine. Parents of K-12 kids, my life as an engineer (travel to a remote site with a "go home and pack now; your flight is in 2 hours" kind of notice). And my life now, where to some degree my daily schedule depends on the weather (it isn't raining so i go do landscaping).

And I'll interject two of my known relevant architectural biases here:

1. As someone who has trained under an Olympic nominee, an Olympic coach and national-level leaders in their field, I understand that part of mental discipline is the athlete never ignores the plan. To borrow from the movie Bull Durham, at times the athlete is "meat"; don't think, we don't pay you to think. The plan can be changed in coordination with the coach, but commitment and mental discipline does not allow ignoring the plan. I think this is ultra critical for building muscle volume and strength because I hear Dr. Isreatel talk about consistency being one of the top key factors. But, once the athlete embraces the idea that they can flex the plan as they see fit, then pretty soon the least little hang-nail becomes an excuse to change the plan. Some athletes will back off below minimum and some will push too hard. It's the coaches job to know what's right and the coach's knowledge is written in the plan. "So let it be written, so let it be done."
2. I view apps like Dr. M. in their achievable ideal form as "assistants", broadly defined to include coaches. There are a lot of advantages to have a digital assistant in our pockets (vs flesh and blood). Availability and costs being the big ones. There are serious drawbacks and it's beyond the scope of this writing to list them all. Relevant to this conversation though is that apps lack "understanding" (the human to human kind) and "empathy" (I know you're tired and everything hurts, but this is a strength phase so we need to push through). I guess a third one is "context" in that an in-person coach knows when to give you a hug and push you in to training, and when to scream and show their disgust at your moral weakness, shaming you in to training. The punchline of this bias is I strongly believe the future of these apps is to transition from "software on my phone" to "assistants" to "coaches", and that good apps are already moving to become assistants.

Dr. M scheduling app (screenshot attached). First, I have to tell Dr. M. which days I want to work out, and that answer is "fixed" forever. Your reply might be it isn't fixed forever; you can change it any time. But that's not correct because any change I make today for this week's schedule becomes all of the following weeks' schedule. A counter might be to change it every week. My answer would be that I lack the information necessary to create a decent immediate schedule without the coach's input. The result is I ignore Dr. M when it comes to scheduling and only use it for programming.

As might be Dr. M scheduling and creating the workout plan:

• I tell Dr. M. what my fitness goals are. First, I need to lose belly fat and get my %fat down to about 20%, while preserving as much muscle as possible. Second, I need to increase cardiovascular fitness, particularly heartrate variability and VO2max. In a perfect world there is discussion on how hard I'm willing to push, any health restrictions (bum right foot so any impact activities are "out"). These types of inputs are critical for creating a training plan. Dr. M. does part of this today, and maybe it's not in the strategic roadmap to grow beyond a weight-lifting app.
• I create a baseline schedule by telling it which days I'm available and which I'm not. Catholics might not exercise on Sunday whereas Muslims might not on Friday. Some may have a club or other social event one night every week (ex book club on Tuesday evenings). Whatever my reasons, that's my "available to work out" days and propagates forward until I change it. I think Dr. M does this today. But not every night ... say five nights in a row ... should be workout days if Dr. M. has me on an upper/lower split. IDK if Dr. M accounts for that.
• Occasionally (probably periodically, like Sunday evening for me) I adjust the next seven days with the reality of my other schedule commitments. Ex, my kid is in varsity football and there is a home game on Wednesday one hour after work. My boss and his wife are coming to dinner on Tuesday and I need to be home early to prepare. I should be able to black out those specific dates in Dr. M without changing the baseline schedule. This is where Dr. M. falls down; the schedule app doesn't allow (that I'm aware of) an change to just the next few days).
• Next scheduled workout planning: Dr. M. needs to understand WHY I'm blacking out those days, at least in a binary sense. If I block out the next four days to take a four-day weekend of drunken debaucher, then that is one type of adjustment to day 5's workout. As opposed to I'm blocking out the next four days to build a stack-block retaining wall or hike 50 miles in the mountains, then day 5 might best be a day of rest. Your answer might be it's up to me to make those adjustments, and my reply would be "see bias 1, I am meat."
• Multi-week schedule planning: I can extend this last point to forward-planning. (For reference, I'm a student of Bompa's periodization (yea, I'm old) and Renaissance Periodization when the "app" was an Excel worksheet). Let's say my long term goals dictate I do a hypertrophy cycle, and I've been lifting long enough that I'm through anatomical adaption and recovery. RP's hypertrophy is (from memory) five weeks: four weeks of building and one week of consolidation. If I told an ideal app assistant that I'd be on a 5 day cruise during week 3, then there's no reason to start the hypertrophy cycle now. Instead maybe continue consolidation with a mini-cut. Do the cruise. THEN on the week I'm back from the cruise (week 4) do one week of adaptation, and START the hypertrophy cycle on the following week (week 5).

I know I have talked about things Dr. M doesn't advertise, like cardio fitness. And while I personally am looking for an integrated fitness assistant that has excellent strength and hypertrophy programming, I recognized Dr. M doesn't advertise itself as general fitness app. I think these comments are valid because all humans are more than just weight lifting, and a strength and hypertrophy "assistant" needs to be aware of how that changes the programming.

NEXT STEPS:

The inclusion of a two-tier schedule (baseline vs "this week") seems straight-forward implementation IF the architects buy in to the model of human behavior that I describe above.

The inclusion of recognizing that some missed gym days are actually good for health and muscle growth, and modifying how the app responds to that knowledge for the next workout, is a bit harder programming but still only relies on buying in to the description above.

The inclusion of multi-week forward planning modifications likely require extensive discussions with experts in weight training (i.e. my story above might not be the correct response, even for a human coach). It also requires an occasional (periodic?) effort by Dr. M. to re-baseline the athletes goals, progress against those goals, and the effort they're NOW willing to put forth. These seems a fairly extensive modification to the app, as well as a few architecture flow-diagrams to make sure we understand how the athlete works and how the app should respond and when.

Lastly, the inclusion of a knowledgeable and empathetic "coach" A.I. is probably a serious stretch goal. IDK of anyone working on that level of sensitivity other than companion-bots.

That was fun couple of hours. Hope it provides some grist for the mill.

Bruce

Would the AI be able to change the sequence to get variety?

I'd love to be able to disable all AI chat integration. For a myriad of reasons I'm morally opposed to it, and would really prefer to just have it off and stick to the workout planning.

Workout demos shouldn’t be links to YouTube. 
i understand the convenience of it but we can all use YouTube if we want. In app, it should be a nice gif or video demo of the exercise. This cuts not on the apps need for internet access. I’m in the gym Train, airplane mode is kind of mandatory. 

It would be a good idea to have somewhere that tells you how long a certain workout. this will help if you are a bit pushed for time sometimes.

The addition of integrated gps would be very useful for distance based workouts. Even the abitlity to have distance based workouts and enter distances manually.

The GPS feature would be for any exercises that could or should be measured by distance. For instance if you go for a run outside, instead of measuring the exercise in seconds (time), you could measure the exercise by distance. The future would require that you start the exercise and the GPS would run in the background, measuring the distance until you stop the exercise. Progress for such exercises would be measured in distance instead of time or reps. An alternative to that for this program would be to be able to use a third party GPS to determine distance and then enter it into this app manually. It just does not make much sense to measure a run or a walk or any other similar traveling exercise in seconds or reps.

I hope that clarifies my suggestion a little bit.

Having to move to, pick up and tap the phone every set is a bit clumsy.  I thought it would  be nice to have a button that would trigger the default action (next/finished).  Something like THIS.  The app would need to recognize the device, map to the appropriate logical button and respond within the current state of the application. 

Add an estimation of calories burnt after workout. Its been suggested quite some time ago so reposting the idea