Opinions needed. Create a Bodybuilding AI for UGM

Status
Not open for further replies.

TG1

Staff member
So for those that aren’t tech savvy. There’s a Chatgpt where you can ask questions and it gives you good answers. But the artificial intelligence isn’t designed particularly for bodybuilding or fitness. It will give you general responses and stays away from exact information. For now it’s basic

image


It’s possible to “train” the intelligence to focus on bodybuilding and fitness. It’s possible to ask it to create you a cycle based on what you have available and it would do so. Give you exact workouts like a trainer would.

It wouldn’t be fast nor cheap to accomplish. I’ve already done some front end work to connect to Davinci from OpenAI. But the backend training and data sourcing will be awhile and cost. It would entail taking data from many bodybuilding books and programs and books on anabolics and cycles and train the artificial intelligence to respond firstly based on the data in the books. It’s very data entry intensive to do.

That being said it would serve a much wider community more accurately then a single person or trainer could accomplish. But is it worth the cost to setup. I can’t really estimate the cost on a small scale. As I would think you would need 3-4 people full time working on the data and training the AI for a few months.
  • Build It
  • Not worth the time
0 voters
@UGMain @Sponsor_G @staff
 
Not entirely. It wouldn’t get rid of contest prep coach’s or training for more advanced levels where you need to compare and see where people need to bring up areas and work on. At least not yet. I think it will be coming when some gym chain or startup focuses on a full training model that will work off using other sensors to better assist.

But they won’t focus on the cycles and what it takes on that side.
 
Doing some checking I see there’s a group of doctors making on for physical therapy already. To include scans of the person doing the exercises to make sure they are doing correctly. So the AI would watch thru the camera and report where to adjust or direct the proper movement.
 
The problem with this is training data: most dudes and even professionals have problems recommending cycles. Also, the AI won’t have access to things like the user’s bloodwork or medical history. If these challenges can be addressed, it might be good. The problem is that you would have to be very careful what you use as training material. Good information on hypertrophy, strength and nutrition are pretty easy to find, but good information on ped usage with an eye towards harm reduction is practically nonexistent
 
Yeah good points. It’s possible to take the users info into consideration but that would really be way beyond the amount of time and training I’d be able to handle. That would be a legit product on its own with investors to cover. Well beyond this at this point.
 
Thats wild ive always been amazed by people who can do that stuff. Ive got nephews taking coding classes in elementary school which idk what that even means but seems crazy when we were fighting and shooting and playing in the woods growing up
 
I would give a small donation 💪,but i know more than any A.I. When it comes to the iron game. Wink / wink what i have might use for is how far does this VR, AR go with SEX. Can i put a 35 yo wife pic and make it like the old days. No more pulled muscles(haha) or arguments about whos on top! Its not that disgusting @John to think about…dont start. Before andro and menopause and complaining! The days of Z cavarircci, v necks…the nautilus double chest!🤔🤔🤔
 
Status
Not open for further replies.
Back
Top