Knowing how NOT to do something, does not imply that you know how to do it.
You can learn either by trial and error or by following the wise advice (imitating) of someone. I have already made use of evolution to explain how cultural knowledge works. If you opt to use somebody’s experience you are piggybacking in his cultural knowledge, which is passed on to you. Since I will be writing about learning, it felt appropriate to use evolution again. In fact, Tim Harford did it for me in his fantastic TedTalk
Well, guess what? During my PhD I discovered that trial and error sucks. In fact it sucks when:
a) you have had too many errors and no win.
b) you could have saved time by following some wise advice from above, but above thought it was better for you to go through trial and error.
As a matter of fact, it is useful to learn on time that it will be beneficial for you to get small results soon to boost your morale so you avoid the PhD dip. It is also good to be aware that in science you need a network and that you should cultivate it in your PhD. Why asking for a scholarship to attend an international conference when you can stay at the uni and focus on your work? Maybe knowing how to convince people to sponsor you won’t be useful when you have to write grant proposals to continue with your research.
Many of these things fall in the category “I wish I had started doing them earlier” or “That sour mouth tast of wasted time”.