Go Fit

@tombielecki Artificial intelligence allows you to completely replace your personal trainer. Workouts are selected based on the physical condition of the body, goals and contraindications. Artificial intelligence will select an analogue of the exercise for users with a contraindication to the exercise or the load on certain muscle groups. Also, there is the opportunity to synchronize exercises with the menstrual cycle of women. Another advantage of Artificial Intelligence is that it can monitor the user's state in the course of the training, adjust the weight and the number of repeats depending on the frequency of the heartbeat. Thanks for your question!

@tombielecki Fitbod - to increase muscles, it is a question of serious sport. We are about the lifestyle of regular people. Our goal is for people to training more often. We are not a gym, we are about playing sports everywhere and always at any request

@dulitharw Thanks for the idea! We already have this one in our roadmap for the latest releases ;)

@rueter Hey dude! Thanks for your question✌️The thing is that now we’re hardly working on the technology itself and unfortunately with very limited resources. Soon we’re going to make landing more detailed and will optimize for sure😜

@lukasz_krawczuk we are preparing something special that will help you :) announce will be in a couple of days, stay in touch

@lyondhur Of course we will have the programs approved by certified fitness coaches. But every user assumes all risks of injury by agreeing the terms of use. And we will suggest to consult with physician before starting our or any other fitness program. This is the number one rule for any fitness apps.

@alexmalkovich perhaps I’m missing the powered by Artificial Intelligence component then. If the programs themselves are certified (powered) by human coaches — matching parameters, bio types and intensity levels — AI in this case would then..?

@lyondhur Artificial Intelligence does not generate programs. Its task is to give the user exactly those programs that are suitable for his purposes, physical fitness and health. In addition, based on the collected data, we can ensure that the number of repetitions and weight for exercises is chosen correctly, find weaknesses and increase the load on them, monitor the correct implementation of certain exercises. 👌

@alexmalkovich I still see some liability mate. I have worked on a project with similar implications - not exercise prescriptions - and there were just a lot of work around that. For example, in any pre-trained exercise model, a bad set of captured data and/or bio-inputs (due to malfunction of the fitness tracker, the HRM or the smartwatch, for instance) could translate into an incorrect prescription with the potential for injury or over training. I understand why you might not want to call it “prescription” for the recommendations provided by the AI embedded in the app. However, think of the certified programs as ingredients and the AI as a chef/cook. The chef can cook a multitude of recipes with the ingredients it is given. It is. 1/2

@alexmalkovich In the unlikely event of such a problem, I reckon you’ll have a horribly bad time trying to justify an incorrect exercise prescription stating that the quality of the data from the capturing devices/gadgets was not good. The only reason I commented on this project is because not only it’s a part of my job I take great interest at, but also because I would be stoked to see you guys succeed with something fresh — amidst a sea of sameness. There might be something here for you guys to look at, if I may say so. Cheers for the response Alex. 2/2

@chrismessina Hi! At start we will use unsupervised learning. We will generate the dataset by analyzing users😊

@alexmalkovich wow...! I see. So are you using pretrained models then?

@chrismessina You are absolutely right! 😊