So, he watches the YouTube video outside the room, and I remind him that if he felt resistance as he was pushing the needle in, that was probably bone, and he'd need to withdraw and try again. So, he starts on the patient (who is awake, but out of it) and pushes in hard. He keeps pushing, and finally says, "I don't think I got it in the right place." So I go to pull the needle out of the patient's spine, and I can't. The needle's stuck in there. When we finally got it twisted out of the patient's spinal column, we saw the damn needle was bent at a 45-degree angle because he'd slammed it into the bone so hard. The patient never knew (drugs erase all mistakes and bleach every sin).
"Why was he practicing on a live human being?" you might reasonably ask. The answer is that there is no perfect analog for a live patient, and somebody had to be his first. That's why we have teaching hospitals: you can't learn everything with dummies and simulations. We're all OK with this in theory, but in practice it means at some point you'll be at the hospital and someone will be training on you.
It might be a nurse doing an IV, a physical therapist getting you out of bed, or it might be your doctor. In the good old days, medical students got more of this hands-on training before they graduated, but due to changes in medical education, brand-new doctors often have to learn on the job. As a senior resident, I've walked interns through everything from a pelvic exam to a lumbar puncture to a central line placement (that last one might not sound too bad, until you realize a "central line" is a large IV usually inserted directly into your jugular).