Show HN: Weight Lifting Algorithm Based on 5M Workouts

https://www.getfitnessai.com/

by jakemor • 7 years ago

70 90 7 years ago

jakemor28• 7 years ago

Fellow hackers,

After collecting an insane amount of data with my app Lift Log, I decided to design an algorithm that optimizes for accelerated muscle growth over a short period of time.

It currently outperform human trainers by 1-2x in the first 60 days of training, using a very simple algorithm. The opportunity to change how people, athletes, and soldiers train is huge.

We're cashflow positive, raised a small seed, and are hiring data scientists & engineers in NYC to hack the human body for efficiency.

Hope to answer any questions you may have!

_Schizotypy• 7 years ago

What's your evidence for this "outperforms human trainers" claim? I only see one single blog post and not much else on the website.

jakemor28• 7 years ago

Sorry for the confusion — website is still new.

In analyzing people's percent change in total volume lifted over 60 days on FitnessAI vs users on Lift Log (a basic weightlifting journal app)

jdotjdot• 7 years ago

This looks exciting, but as other commenters noted, you need to provide more detail about the underlying data and ML strategy. There are many companies peddling both questionable "AI" and questionable "sure-to-work" training and fitness practices, and you sit in the intersection of both. Even if what you offer is real and truly works (and I hope it does!), you'll need to show a bit more to get people to trust you and invest their time in following FitnessAI's instructions.

jakemor28• 7 years ago

Noted! Thanks so much for your input — I'll work on a post

dyeje• 7 years ago

What exactly does the algorithm optimize for? I see you mention muscle growth, but does this come at the expense of other factors such as preventing injuries?

jakemor28• 7 years ago

% change in total volume lifted over time

superpermutat0r• 7 years ago

I can easily outperform your app on any client by assigning more sets per during the week, more tonnage, than the app. It's the fine line of not getting adapted to bunch of work in the beginning to enable long term gains. People should be doing the minimal necessary dosage of stimulus.

eranation• 7 years ago

If a user has some constraints such as lack of some equipments or some movements they can’t do due to health reasons, do you guys handle it? It would be amazing if your algorithms can work around those constraints (if there are some exercises the user can’t do, offer an alternative based on what the user can do and what equipment they have access to). I didn’t see it in the app.

Also I know that everyone does it, but the “free trial but we’ll charge you even if you don’t use the app and just forgot to cancel” is something that I really dislike. Maybe it’s an AppStore limitation but I rather do a trial and only pay if I actually use it, not if I forgot to cancel.

Lastly, maybe it’s after the paywall, but not everyone knows what all the routine names mean, if I have to google it, it’s something I expect the app to include. (A diagram / video of how to do it the right way)

Other than that, looks exciting and a really nice take on this topic.

jakemor28• 7 years ago

Hey! Message me in app for a free month :)

You can modify the routine but not based on available equipment and we have full step by step instructions along with videos for each exercise

Regarding the subscription thing — that's an app store limitation

Stunting5• 7 years ago

I'm all here for better lifting apps, but I have lots of questions.

How do you qualify the "trainers," that your algorithm outperformed? Most basic strength and conditioning books will outperform the average chain gym trainer.

How many of your the trainers in your sample size were Strength and Conditioning Coaches at reputable performance gyms? Would any qualify under the following statement from the National Strength and Conditioning Association's Website:

"The proper education consists of an undergraduate degree in an exercise science-related field. Such a degree is imperative because it forms the foundation of knowledge that underpins the profession. An exercise science-related degree affords the capability to practice an evidence-based approach to strength and conditioning."

What are your results for longer than 60 days? Do you continue to outperform human trainers at the 1 yr and 5 yr marks? For those that are uninitiated, 60 days is a incredibly short time in the world of Strength and Conditioning. Doing almost anything differently that your normal routine, consistently, over 3 months will yield positive results before plateaus set in.

How are you quantifying your results?

Are you raising the 1 rep maxes of your athletes 1-2x than more than Human Trainer? Their 10 rep max? Their max broad jump? Their 40 yard dash? Their 1 mile time? Their muscle mass? Lowering their Body fat percentage?

I want this to work. I crave a simple logging app that will allow me to enter a variety of exercises and an AI that would help with programming would be huge. I would be willing to you pay your subscription of $8 a month if it were true.

jakemor28• 7 years ago

> How do you qualify the "trainers," that your algorithm outperformed? Most basic strength and conditioning books will outperform the average chain gym trainer.

Trainers use Lift Log to keep track of their clients workouts during training sessions. They were identified and their data was compared

> How many of your the trainers in your sample size were Strength and Conditioning Coaches at reputable performance gyms? Would any qualify under the following statement from the National Strength and Conditioning Association's Website:

Not sure! but they worked at or owned their own gyms so I'm sure they are certified. Only trainers who trained using the same exercises as FitnessAI were compared

> What are your results for longer than 60 days? Do you continue to outperform human trainers at the 1 yr and 5 yr marks? For those that are uninitiated, 60 days is a incredibly short time in the world of Strength and Conditioning. Doing almost anything differently that your normal routine, consistently, over 3 months will yield positive results before plateaus set in.

I'm not sure! We've only been around for a few months so it's hard to say. That said, we have the data and users to optimize for this once users begin to plateau

> How are you quantifying your results? Are you raising the 1 rep maxes of your athletes 1-2x than more than Human Trainer? Their 10 rep max? Their max broad jump? Their 40 yard dash? Their 1 mile time? Their muscle mass? Lowering their Body fat percentage?

% change in total volume lifter

> I want this to work. I crave a simple logging app that will allow me to enter a variety of exercises and an AI that would help with programming would be huge. I would be willing to you pay your subscription of $8 a month if it were true.

try it out! And message me in app for a free 30 days

halbritt4• 7 years ago

I can outperform 90% of human trainers as well in the first 60 days of training as most human trainers are clueless and the novice effect is real.

Do any compound movement with as much weight as you can manage for a set of 5. Do a couple more sets. Add 10 pounds to the bar every subsequent workout until failure. Then add 5 pounds to the bar every workout subsequent to that.

That'll get most healthy young people through the first 8 weeks of training. Train two or three times a week and you'll add 80-100 pounds to every lift.

The caveat being that small people will run out of linear gains in their upper body during that time.

jakemor28• 7 years ago

> Do any compound movement with as much weight as you can manage for a set of 5. Do a couple more sets. Add 10 pounds to the bar every subsequent workout until failure. Then add 5 pounds to the bar every workout subsequent to that.

The app basically does this very systematically. Give it a shot! It's simple and effective

Obi_Juan_Kenobi2• 7 years ago

60 days isn't long enough to evaluate program performance; that's about the length of one mesocycle and thus it is not sufficient for evaluating how that performance carries forward. E.g. you could compare a peaking cycle to a volume cycle and falsely conclude that the peaking cycle is strictly better for strength gains. However, peaking is not sustainable and only provides temporary adaptations.

You're also analyzing increases in volume, whereas most people will be interested in increasing 1RM strength. Volume is obviously a key component of this, but volume as a metric is also tricky. The number of challenging sets is likely a better indicator of training stress and a driver of muscle growth.

https://www.strongerbyscience.com/can-we-predict-muscle-grow...

For instance, it's trivial to increase total volume by performing high reps of a light weight over a long period of time. I can spend all day lifting 35% of my 1RM, accumulate astronomical volume, but see little to no training benefit.

I also take issue with your claim that your algorithm 'optimizes for accelerated muscle growth', yet it does not appear you have measured muscle growth at all. It's an absurd claim to make; you've optimized for a setsXrepsXweight volume metric. This is perhaps a semantic issue, but deceptive nonetheless.

______

I imagine that the app results in reasonable programming, and it may even be effective, but you have a long way to go before claiming any kind of empirical validation.

jayess• 7 years ago

Android version? I'd love to use this.

jakemor28• 7 years ago

Top priority — if rev stays steady and we can afford it absolutely!

BenMorganIO• 7 years ago

Also an android user, also interested :)

z5h• 7 years ago

You want my phone number AND "This app does use third party services that may collect information used to identify you."

I'm not so comfortable with that. Can you tell us more?

the_duke• 7 years ago

Accidentally, this is not GDPR compliant, they need a list with all third party services.

jakemor28• 7 years ago

Hey! We use the Facebook SDK & AppsFlyer to help optimize + attribute marketing campaigns. Also use Intercom for customer support.

shapiro92• 7 years ago

is there ANY science behind this? did you take into account weight, height, weather, diet, mood, sleep time, gender, prior experience, age, stress levels, drug consumption, proteins... I could go on..

jakemor28• 7 years ago

Hey! Its a pretty naive algorithm that tries to increment your weight lifting goals similarly to how the best of the best did in the dataset.

amelius• 7 years ago

Also, how big is the improvement? If it's only 1-10% better than doing a random program, then perhaps it's not really worth the trouble. However, if it's 25+ percent more effective, then we're talking.

(Of course, for professional bodybuilders the decision may be at a different point)

nolite• 7 years ago

How did you get 6,000,000 workout datasets from 30,000 people?

jakemor28• 7 years ago

A previous app I made called Lift Log for iOS that has been around for 3 years now

danmg• 7 years ago

Are you giving those users free access to this service, since you are profiting off their data?

smn1234• 7 years ago

I'm not entirely sold.

Fitness magazines assert they can get me bigger, faster in every month's issue. It's a big industry.

Perhaps you can post a bit more on the Blog to establish credibility.

jakemor28• 7 years ago

Message me I'll give you a free month :)

Working on blog posts too

aksels• 7 years ago

my 2c: The default value for "weight" in onboarding seems to be 120kg. I'm pretty sure you can set it to 75kg[1] to avoid a lot of scrolling in the onboarding.

However good job !

[1] average: https://en.wikipedia.org/wiki/Human_body_weight

jakemor28• 7 years ago

Thank you!!!

It's actually set to 120 for both kg and lbs — my bad!

rlander• 7 years ago

This seems targeted at bodybuilding (at least the landing page screenshots are). If this really works, it would be interesting to see how the algorithm fares in competitive sports like olympic weightlifting and triathlon.

jppope2• 7 years ago

Seems odd that they are only supporting IOS. Additionally, who knows if they got it right without supplying any supplementary documentation...

jakemor28• 7 years ago

Sorry! I'm a 1 man show and iOS is my specialty

jppope2• 7 years ago

well that explains it! looks like a cool project!

superpie• 7 years ago

Not a fan of the predatory "Sign up for a 7 day free trial, and if you forget to cancel we'll bill you $60 for the year membership" system. It's also a week trial, but the AI takes 8 workouts to learn -- not great for a comprehensive demo at all.

jakemor28• 7 years ago

Sorry! message me in app I'll give you a free month

(stands for anyone from HN)

komali24• 7 years ago

Ok cool but... Why not just not do that? Is the one month of revenue from people who forget to cancel, then cancel and are pissed about it, Worth it?

nerbert• 7 years ago

I don't think it's really the point made there.

throwaway2019X2• 7 years ago

Message you.. in app...

In the same app that he's reluctant to sign up for in the first place because of the predatory trial.

starpilot• 7 years ago

At least with iOS, you can cancel the trial immediately after signup, and your account is still active until trial period ends.

ds0• 7 years ago

I'm wondering if there's a parallel for nutrition. I'm a pretty poor eater, as it stands.

deagle502• 7 years ago

Their blog post says not to rest for more than 90 seconds between sets. I can see why the app would work better with tight groupings but 90s is not enough max gains and recovery, especially for beginners.

jakemor28• 7 years ago

Consistency trumps all! (according to the data)

komali24• 7 years ago

90s Jesus fuck I'd die. 2:30 minimum for me, usually more like 3:00.

halbritt4• 7 years ago

If you're doing heavy compound lifts like squat and deadlift, and you care about developing strength, then take as much time as you need to succeed on the next set.

swat535• 7 years ago

on higher weights sometimes i wait a full 5 minutes. from my experience resting more ensures you complete all your sets and reps which are the most important things based on my experience.

at a minimum i wait until my heart rate slows down. i haven't noticed any additional gains for shorter weight times.

Afton2• 7 years ago

Beginners in the first month or two are mostly just learning the motions. 90s rest is fine. Once you're a few months in, and you've progressed all the weights significantly from your untrained starting point, then 90s seems like a crazy short amount of time.

deagle502• 7 years ago

"The men who took longer rest periods experienced a 152% increase in MPS in the early part of recovery compared to the men who took short-rest intervals who only saw a 76% increase. That’s a two-fold difference."

https://www.mensjournal.com/health-fitness/perfect-amount-re...

Why would you deprive a beginner of these benefits?

halbritt4• 7 years ago

Beginners will be enjoying the "novice effect" and will benefit from handling as much weight as possible after learning the movements, which should happen in the first one or two workouts. Extending the rest periods as long as needed to finish the next set will benefit the novice more than the experienced lifter.

Shortened rest periods are good for prioritizing growth over strength, i.e. body-building. One can only begin "body-building" after establishing a sufficient base of strength from which to start.

HAL9000Ti• 7 years ago

Uh. too bad I don't own a phone

insickness• 7 years ago

Even for a single exercise, there are a lot of different variables: weight, number of sets, number of reps, recovery time between sets, recovery time between workouts, etc. Were all of these variables adjusted and taken into account or were some of them held constant?

bewatson• 7 years ago

I agree, and just as important as weight/reps and number is the speed of reps. Muscle growth comes from the amount of time the muscles are being utilized under weight. If someone cranks out 10 bench press reps quickly, that will have vastly different results than a 3110 lift. (3 seconds up, 1 second hold, 1 second down)

jakemor28• 7 years ago

Weight, number of sets, reps and recovery time were all taken into account, although they weren't necessarily constant.

You can devise a relationship between someone's 1 rep - 12 rep max and adjust for it, check out epleys formula https://en.wikipedia.org/wiki/One-repetition_maximum

brandonmenc• 7 years ago

I see you have a post-set rating for difficulty but "reps in reserve" is also a common assessment - any plans to add it?

jakemor28• 7 years ago

Yes in due time!

Stunting5• 7 years ago

Jakemor, I noticed you haven’t responded to anybody asking you more than a generic question about your methodology.

Here’s what it looks like you did;

Made a weight lifting logging app and charged a monthly subscription for it.

Used your users data to create an algorithm that culled the routine of your best performers, most likely defined by heaviest weight lifted.

Claimed that by following your algorithm your users could 2x the results of trainers; trainers here is defined as “people who train,” And not “people who are paid to train others.”

Cited 60 days as a window of success.

All of this, if it is as it appears, is massively misleading in an industry already known for snake oil.

Have you compensated for your users that had big numbers but approached it in an unhealthy way? For instance, how many of your user’s deadlift high numbers resulted in a hernias?

Additionally you say you have engineers and scientists working on hacking human performance. What about strength and conditioning coaches? What about athletes?

throwaway2019X2• 7 years ago

> Additionally you say you have engineers and scientists working on hacking human performance.

In this post, he says he's a 1 man shop.

https://news.ycombinator.com/item?id=19635573

jakemor28• 7 years ago

I said I'm looking to hire scientists and engineers

jakemor28• 7 years ago

Hey sorry to seem misleading — that isn't my intention.

I created an algorithm that trains you based on how top performers on Lift Log train themselves.

Top performers = highest percent change in total volume lifted from day 1. Sure, 60 days isn't a ton of data, but it is enough to show a difference.

Most people don't train properly and don't increase the difficulty of their exercises between workouts — following FitnessAI's algorithm (although it is incredibly naive) does indeed result in the performance boosts claimed. Try it out!

Regarding comparing to trainers — these are real trainers. I identified many trainers using lift log (on their clients phone) to keep track of training sessions.

pushcx• 7 years ago

FitnessAI is so simple your grandma can use it.

This is a terrible cliche. (More: https://geekfeminism.wikia.org/wiki/So_simple,_your_mother_c...)

How about something like:

* Your lifts should be hard, your tracking should be effortless. Record your set with a single tap and spend your rest recovering for the next (...or taking selfies, we don't judge).

This tells people exactly what the value is and tells the that the user that they deserve FitnessAI.

jakemor28• 7 years ago

When you're right, you're right! Thanks for the rec. I've updated the site :)

ksaj• 7 years ago

Good call for both of you. I'd be afraid to insult this grandma's intelligence when it comes to working out: https://www.facebook.com/uniladmag/videos/72-year-old-woman-...

PS: Sorry for Facebook and sorry for really awful music. Hit mute first.

papermachete• 7 years ago

Awesome! Where's the source code?

agentdrtran• 7 years ago

This is complete snake oil.

perrohunter• 7 years ago

I came to the comments expecting a lot skeptics and was nos let down. Guys give this a try, I fit the usual software engineer physical stereotype and would like to gain some muscle but I don’t invest that much on PT as i should.

IMO, Give this a try jakemor said in the comments he’ll give a free month to anyone on HN

Stunting5• 7 years ago

Hey bud. Instead of sending $8 a month to have information that is already accessible to you regurgitated to you with pretty UI try this instead...

Do you walk? If you don't, start there.

A study of sedentary, overweight men and women (aged 40 to 65 years) showed they lost body fat and weight when they walked or ran 12 miles a week during an 8-month study, without changing their diet. A control group of non-exercisers all gained weight and fat during the 8-month study."

https://jamanetwork.com/journals/jamainternalmedicine/fullar...

Do you sit at a desk a lot? You probably have poor posture associated with it. Do any yoga, at all. Literally any program.

Here's one from my favorite online yogi - https://www.youtube.com/watch?v=tAUf7aajBWE

Do you want to lift weights? For $8, one time, you can order Starting Strength by Mark Rippetoe. - https://www.amazon.com/Starting-Strength-Basic-Barbell-Train...

This book will teach you how to lift the weights, as well as how to program the lifts, which is twice as much as this app is claiming to do. At 1/12th the price for the first year, 1/24th the second, and 1/36th the third, etc..

Obi_Juan_Kenobi2• 7 years ago

Anyone that puts reasonable effort into a reasonable program will see great results.

Definitely give something a try, but I see no compelling reason to pay for this when there are a variety of high-quality free programs out there.

The importance of 'fitness programming' per se is massively inflated online; you'll absolutely need some structure to achieve high performance, but 'the usual software engineer physical stereotype' basically can't help but make progress if they're actually doing physical activity. You could do some of the most bone-headed things imaginable and you wouldn't be wasting your time.

If you're looking for tracking apps or something to do the 'thinking' for you, there are literally dozens of good options.

corecore• 7 years ago

As someone who lifts 5-6 times a week, let me give you a solid blueprint based on what has worked for me, my peers, and many others who know what the fuck they're doing.

It's important to note that this will be a process and anyone trying to sell you a shortcut is full of shit. Including the guy who made this thread. Cool idea, but misleading as shit.

Here we go.

1. Your diet will fuel your growth way more than any workout routine generated by a man, woman, alien, genderless entity, or artificial intelligence ever will.

To put it simply, if you want to pack on muscle? Weight train and eat eat more calories than you burn. ~300-500 more calories.

If you want to lose weight? Weight train and/or cardio. Eat less calories than you burn. ~300-500 less calories.

You can find many TDEE calculators online. I like to use the calculator over at iifym.com.

For example:

If I want to gain weight and my body burns around 2000 calories per day, then I should probably eat 2300-2500 calories, with my protein intake being around 0.7g/lb of body weight.

If I want to lose weight and my body burns around then maybe I should eat around 1500-1700 calories, with my protein intake being around 0.7g/lb of body weight.

That's it. It's a numbers game. Calories in vs calories out. Your body doesn't give a shit about anything else (assuming you're eating balanced meals to fulfill your vitamin needs. If not, pick up some multivitamin and you'll be fine). The hard part is usually planning the meals and actually getting used to eating a lot. That's going to suck, but it gets easier.

And forget all that shit about supplements. Only thing you truly will find useful is whey protein powders and creatine. Do your own research. You're a smart guy, I bet. You got this.

2. Your routine is secondary to your diet. Remember this shit. Don't forget this shit. Write this shit down. Your routine is secondary to your diet. Aside from that, you need to know what you want to do.

Do you want to be strong as fuck? Find a routine that focuses on strength training.

Do you want to be aesthetic as fuck? Find a routine that focuses on hypertrophy. This will most likely be a program that works each body part twice a week.

No matter what you decide to do, be patient and consistent. Results will come, but never overnight.

Whatever routine you choose to do, focus on progressive overloading aka increase the stress put on your muscles each workout. This could be by increasing the number of sets, reps, or weight. Just do more.

Find a routine. Make sure it has a focus on compound exercises. Find a personal trainer who can actually show you how to lift correctly (should only be a few sessions). Go hard. Be consistent. Don't switch the program until you at least put in 6 months of work or your goals change. Don't be afraid to tweak things to your own liking.

shawnvo• 7 years ago

There seems to be a lot of shade here and I wanted to offer a different perspective. I've strength training for over 10 years and aside from the basics like nutrition, rest, form, etc... one of the most valuable tools I use are progression spreadsheets like this one: https://docs.google.com/spreadsheets/d/1VYNuMhonQw1SOKO_7IbG...

The main value is that it not only tells you what exercises to do, but also how much more to lift each day to ensure steady growth. They are designed by well-informed people with a lot of experience in the field. However, there are some major flaws (e.g Most are just linear increases, they don't handle plateaus/failures well, they aren't personalized, exercises don't change, and the UI is pretty bad).

There's a lot of potential in something like fitnessai to address these issues. Obviously, it's not going to do everything for you and won't replace your own research and experimentation, but I can see it being super valuable as a tool in the gym.

@jakemor the app looks dope. Listen to your customers and keep building it out. The majority of HN may not your ideal customer but don't let that deter you.

sawmurai• 7 years ago

I think this will not work unless things like individual training experience, proper form execution and diet are taken into account.

I also think that people are pretty stupid and you will earn a lot of money from it.

ken2• 7 years ago

In various places on the webpage, it brags about the design, and how you can do things with "only a few taps". That'd be great, but the first screenshot shows how I have to "Speak to FitnessAI". Do I really have to have an ELIZA-style chat, and spell out the muscle groups I want to work out today? This just looks lazy. A pseudo-English command line is a terrible design for this.

You have the official "Download on the App Store" badge, but it links to a web form that requests my phone number. That's not an acceptable use of Apple's badge, according to their rules on the webpage you downloaded it from.

Using algorithms to optimize my workouts sounds like a good idea in principle, but I see no proof that it works as advertised, and everything else about this app is mildly distasteful.

jakemor28• 7 years ago

You're only hit with the chat during sign up, after that it's optional. Beginners like the feature.

The download badge links directly to the App Store on iOS; on desktop it lets you send yourself a text with the download link, which is standard behavior for Branch links. Alternatively you can just search FitnessAI in the App Store

I'd suggest giving it a shot before drawing any conclusions. Message me for a free 30 days :)

codesternews• 7 years ago

will someone provide link to app? It's really bad to collect the phone no instead you can ask for emails.

jakemor28• 7 years ago

Here you go! If you visit the site on mobile it links directly. https://itunes.apple.com/us/app/fitness-ai-weight-lifting-pl...

Show HN: Weight Lifting Algorithm Based on 5M Workouts

Related Stories