Okay look, I’m sure everyone’s tired of it, and all the fighting, I mean Nate had to show up to a few podcasts it’s gotten so bad. 
But it’s fundamentally a bad customer experience (for at least some customers). And we can talk about the ends justifying the means, or if people would just go along, or why so angry, but the issue is just the experience of it, and that is not ‘wrong’.
There are as far as I can tell at least 3 major epochs of TR. Chad, Amber, Jacked Nate
The Chad phase is characterized by an epic ton of sweet spot, of ‘hard’ cycling, of pushing your boundaries, of going to extremes. A lot of traditional cycling lore/knowledge blended with a neophyte analysis of the latest scientific studies. For many people this worked, if not making them the fastest at least making the the 'hardest’ and giving a sort of, I was hazed as a kid and made it, so hazing is good, kind of vibe. (And chad is a complex and nuanced human being like all of us, this isn’t a judgement of chad, we love chad, it’s just a vibe).
Then Amber came, is it a reaction to the Chad phase? Who knows, but a pundit might consider that this phase was more science, and more self care. Does this mirror a societal shift toward giving a shit about people’s emotional and physical well-being? well, I mean…one could see the connection. Some shift in the style of content, a shift in the plans, and more science, and perhaps slightly more self consciousness, defensiveness? These are the first days of the ‘tower defense’ podcast. Out of this new ai prediction and someone running actual statistics on the data there’s a realization (this is editorializing by me) that maybe, just MAAAAYBE some people are right about burnout. And to everyone’s credit this new Amber era sees the introduction of levels. This is an attempt to heal the rift. I mean…there are thousands of workouts, all this IP, the history and a (let’s be honest) BANGING app that people really like to use. So levels, shifts to plans, based on actual data. The chad era people are cranky, I was hazed, you should be to.
Then…Jacked Nate. What can be said except, I have a hot tip for you, buy yogurt stock. But the Amber phase is done, podcast is a lean machine at this point, this is the drop the faff stage of the scientific cycle, and push directly into the ‘recipe for success’ phase. Hit your workouts, eat your yogurt, grow a non-ironic stache, and get jacked. This is the rise of the new AI, which is by some accounts an accountant locked in all white room, they have a funny terminal with weird things on it, boxes with names like antelope and phoenix, they have 6 categories with various floats. They have a lot of time series data organized by date, they have a seeming random int with a range between 100 and 500 that changes every 28 days, and they have to manipulate all these boxes to get the best number.
Int comes in, categories are assessed, random int input, future mapping to past, and boop. it spits out a statistical curve, add those together and you have a bell curve of future outcomes.
Welcome to the future. Now the output is just a statistical analysis of the previous two generations of TR, mapped to seeming arbitrary (but highly important) effort gradings. There is no judgment, no insult intended, it’s put the data in, get the answer back. Change one response, and the butterfly wing in Antelope+2 makes your ftp go up 9 watts. But that’s just the statistics. Will that result come to pass? Well there’s a 55/45 chance it will, roll the dice? If you are Chadanian this is anathema, results are the logical outcome of pushing yourself to the limit. If you are Amberite this is strange and somehow dissociated from the emotional safety of I did the workouts as prescribed, if the workouts prescribed aren’t going to help me, then what the hell are you doing?
Three minds, one app. A million complaints on the forum.
I for one, got a 3w drop in my plan, (immediately after getting a new prediction) I changed to balanced it said 9 watt gain. My personal opinion is that it’s confused. I know it’s confused because it was telling me what my ftp was when I knew it was way off, and it tested way off but the app insisted, kept saying no but really its’ this 20w higher number, and now…it’s seeing that it was way off. An artifact of skewed statistical data? Most likely, am I a weird outlier, possible. Do I care? Well I’ll be honest it is a bit…disconcerting, I suppose on some level I’m more an Amberite than a Chadanian, I like the idea of I get a plan, I hit the numbers I gain. But I’ll play the game, what do I have to lose, if it can deliver the 9w then I’m fine trusting our AI overlords (or statistical wonks in all white room).
Oh and actually! I just double checked, moving from demanding -3w, to balanced +17w in 28 days. 20w spread and I can barely tell by tss and hour values what actually changed. Color me curious to see the results.