They did, and then clarified that it wasn’t their claim that coaches would be obsolete but that AI could do the heavy lifting of workout selection.
Coaches would always be advantageous for certain people for accountability, motivation, feedback, skills, etc etc.
I think the crux here is, does intensity management fall into the “workout selection” part or the coaching part. I’d argue that “workout selection” is intended to cover this based on the intent of the feature, so I think this is an issue they seemingly committed to “solve.” However, I’m not naive to think I don’t play a bigger role than some tool in determining my own readiness to do the work, but the claim TR made was a bit more marketing than realistic.
You may be right. To be honest I don’t think personal coaches will ever be made redundant for those with more money, or more specific requirements. Something like AT could very easily take up a part of the work though.
I think there will be more work on plans one day when TR becomes a complete coaching solution. It’s got some great features already for those of us who don’t want to/can’t justify individual coaching. There’s nuance with everything of course.
My point was more about context. Let’s not do that most frustrating of journalistic traits of taking one statement and presenting it as the whole story.
for me, a very average cyclist, one of the primary benefits of having a coach is having that person to manage the amount of intensity in your workouts while prescribing the correct amount of rest and recovery.
Before I had a coach I burned out on a TR half Ironman plan. I think now I have a much better feel for when I need rest and wouldn’t make that mistake again.
But for anybody new to the sport it’s tricky because you see so many posts about intensity and suffering, etc. that you believe you have to suffer during every workout to get better.
For those much faster than me and vying for more podiums I get there’s a fine line between pushing oneself but not too much. For the mere mortals like myself I think the better guidance is to error on too much recovery and focus more on consistency.
And language and meaning and subjectivity plays such a role in all this.
You and I could be the same fitness level, do the same workout, and describe a vastly different experience. Some people like the feeling, some people really feel like its suffering. Does this have an impact on the fitness gains? Maybe not, but it definitely has an impact on motivation.
Then we get to burnout… is it physical or mental? More than likely a combination.
Obviously a good human coach will know these things, but how much are you willing to pay? How much worth is it for you? Do you hire a 500 dollar a month coach to compete in local gravel races and cat 3-4 road races?
Personally that is why I like the TR stuff. And the more I learn and get to know my body, the more I can play a bit more a role as my own coach (though I definitely see the limitations there).
Regarding DJ’s problem with TR plans, right now they seem to work for me. I am on the mid volume using plan builder. 5 workouts a week, plus I typically do an unscheduled ride on Saturday and rest Sunday. It isnt too much intensity YET, and I am still improving. I haven’t dug myself into a hole. Is there an opportunity cost where I could be getting faster with a different type of training? Maybe.
I’m just happy to be living in a time where there is so much information out there. Have to be super careful about where you get it and how you apply it (look at the different opinions and “experts” on this forum alone), but it really is great that the info is there.
I agree. I have a coach who I started with for my first Ironman but liked her so much I’ve stuck with her. But I’m not made of money so there’s definitely a price point I would not go above.
I think there is a lot of good training resources out there from this forum, local cycling groups, podcasts, etc. lots of good guidance. It can just take time to sort through it on your own and learn what works for you.
In advance, sorry for my english:
I think this post is getting out of hand. It started with a topic that was covered in a very clear podcast, then it has fluctuated between different allegations that the volume of the plans is too low or too high, that they have overtrained because of trainerroad, that we are giving away data that they use to profit, someone even proposed a discount.
I think this is simpler: trainerroad is nothing more than a tool that is delivered to us at a certain price (much less than a good coach). That tool includes adaptive training plans of different volumes and also includes hours and hours of podcast and forum pages and pages, which could be considered as the instruction manual of this tool.
No one is forced to use trainerroad (and give them “free data” to improve this tool for us), no one is forced to follow a high volume plan, and no one is forced to read the instructions.
Like any tool, if you don’t know how to take advantage of it, you can end up doing harm (to yourself) or simply be useless.
The trainerroad support team is great, they have always helped me with questions about my training, as has this community. And there are a lot of resources. So there are no excuses to missuse this tool.
It may be safer and even better to have a personal coach, but nobody has ever said otherwise, trainerroad is for people who do not want/can have a coach and it is a tool based on data obtained from a population that is not necessarily valid for everyone, but that doesn’t make it a bad tool… it is a cost effective tool… if you take advantage of your resources… If you don’t take advantage of them or you overestimate yourself and sign up for a high volume plan without even having time to sleep or without following an adequate nutritional plan, and don’t see results or you get overtrained, then the problem is not TR, the problem is you.
Very well stated. Did I hear a mic drop?
Please do.
I base my comment on the announcement of AT, specifically at time stamp 00:03:23:
A little bit later (00:26:10 onwards) clarifies AT’s mission statement a bit more and clarifies the role that TR sees for human coaches in the future (emphasis mine):
I’m pretty sure TR has said similar things in subsequent podcasts, but I don’t want to sleuth through all of them.
Overall, that seems to track exactly with what I stated here: in the future, coaches will use AT and tools like AT to help them select workouts for their athletes. But AT or other ML-based software tools don’t make coaches obsolete just like Photoshop and Illustrator did not make artists obsolete either. They will just shift what they spend their time on to other things.
Personally, I think coaches will be invaluable for things like goal setting, teaching the basics and giving recommendations on e. g. how much volume you should do given your life circumstances, your goals and the like.
Just to embed this into a larger context: if you ask the right people, this is not a new idea. A friend of mine, a math professor and pioneer in machine learning, was offered a consultancy gig by a major league baseball team to automatically analyze the movements of the players. That was in 2014 or 2015, so things have progressed a lot since then. Other experts also told me that sports teams were very interested. (There was a big 3- or 4-month program on machine learning at a well-known math institute, and said friend as well as a few others were among the participants.) Jumbo Visma is using ML to determine the nutrition of its riders in the TdF.
So I don’t think it is really controversial that ML-based techniques like AT will become a commonplace tool for selecting workouts that many coaches will use in the future. I realize I might come across as a TR fanboy, but I love technology and I think @Nate_Pearson has bet on the right technology at the right time (if memory serves, this has been in the works for at least 4–5 years, i. e. just when ML started to become commercially viable; according to @Nate_Pearson, he has had the idea since 2011/2012). Of course, a bet is still a bet, and it doesn’t need to materialize or even seem to have a big impact initially.
As far as I can tell, apart from TR nobody else seems to be doing this for cyclists.
PS I apologize for the messy grammar in the transcript. I tried to hue as closely to the source as possible, and us humans are not very good at grammar in extemporaneous speech.
I don’t think it is reasonable to have that expectation.
In the most recent video of Alex Dowsett he described how he fitted his wife onto a road bike with clip-on TT bars. The video is hilarious (and his wife is super brave). He is a world class TTer, he held the hour record and is still the third-fastest person around the velodrome for an hour. He knows his stuff. When dialing in the fit, he had to make sense of comments like “I feel like a turtle.” and “I feel like a stretched out slug.” and make changes based off of that. 
(Please, nobody think I am making fun of his wife: she is entering his realm as an amateur and he is one of the best in the world. This takes guts. Oh, and then have everything filmed and be put on Youtube.)
What I am trying to say is that his wife didn’t have the experience and the vocabulary to express what was wrong or right with her fit. And the same is true for people who just started structured training or haven’t/don’t want to invest time in learning the basics of training. In my mind, TR at present does a good job accommodating those people — not perfect, there is no such thing, but I think it gets the job done. Just like Dowsett erred on the side of comfort, TR should (and in my experience does) err on the side of ease of use with features like TrainNow.
The second thing that in my mind TR should do is provide an easy on-ramp for people who want to learn more and do more customization. Also here, I think TR does a good job: I find it great that TR tries to educate people, and I can honestly say that I have learnt a metric shirt ton from the people in the forums and listening to the podcast. I wish I was born 20 years later so that I could go through Training Plans in my physical prime without the commitments I have now.
At the other end of the spectrum you have pros or amateurs who compete on the world or national level. Some them use TR, although I reckon many of them will either have a coach or if they are self-coached, then they will roll their own. TR allows that, and even here AT helps to some degree. (PLs and workout levels are a huge help in picking alternates.)
At the risk of bringing up older discussions, this is what I meant with some of trainerroad suggestions. It isn’t that they are wrong, it is that they are erring on the side of caution and trying to find the best compromise that is safest and works for the most people. They are well aware that some people will want to push into more “extreme” areas
What do you mean here?
(I can’t quite tell whether you are adding commentary to my post or are disagreeing with me on something.)
Even most (good) human coaches would (and probably should!) err on the side of caution. That is, unless you have a very close relationship to your coach (e. g. because you are a pro or you have developed a close relationship to them), they know your capabilities and your life circumstances very well, and you both trust each other.
This is one issue of the software-centric approach that TR takes: some people bite off more than they can chew, and/or don’t really trust AT’s recommendations. How do you prevent people overreaching and building trust? It’s hard.
You say its an unreasonable expectation and dont explain why?
Great that you had stories to tell… But if you just want to tell stories why reply to me at all???
I don’t think its reasonable for you to not think that my request is reasonable!
I watched a youtube video tonight on Thor Love and thunder. They were riding around in some sleds pulled by magic goats.
I think I did give explanations. The anecdote of Alex Dowsett fitting TT bars on his wife’s road bike was meant to illustrate my point: existence and lack of knowledge not only shapes one’s perspective, but it can decide what is and isn’t optimal — or even possible.
Feel free to disagree with anything I have said, but I am not quite sure how you can read my post and come away with me not giving any explanations as to why. 
I am not sure where your anger towards me comes from. I don’t mind if we disagree, but I’d like to keep our discussion civil and respectful.
@OreoCookie - @FatBoySlim is saying SBL should be able to give us strength training and stretching workouts within our plans, as many coaching services do. I don’t understand your reply that this is an unreasonable request or how the TT setup example applies to the “complete plans” comment.
If that was his main point, I completely agree with that. IMHO they should also partner with reputable skills coaches to have a library of material for that. Ditto for strength training.
Adding commentary, agreeing, and making a reference to a point I made in a different thread that TR will probably not always suggest the optimal point for fear of customers thinking if “X is good, then +X has to be even better!” when that definitely isn’t the case after a certain point.
I don’t think it’s an unreasonable request so it would be a nice addition to the TR plans. But let’s not pretend it’s a challenge to a TR recommend strength workout.
It took me 5 seconds to google this TR article
https://support.trainerroad.com/hc/en-us/articles/360000087826-Integrating-Strength-Training
So I get it adding it in a plan would be nice but there is no doubt in my mind that if TR added the above plan or some variations to plans there would immediately be complaints about what the workouts are.
Agreed. But there’s a lot more nuance to creating an integrated plan than just naming the exercises and the set count.
Athletica are using machine learning, AI, adaptive training and there’s a coach element now etc.