Training places are already rammed to the gills, there aren't enough places to put substantially more students. Junior doctors already have to compete to get training slots even vaguely where they live and if they miss that they simply have to physically move to a different city. Medical training isn't as simple as just adding a lecture theatre and a few classrooms, or even a whole university building faculty of lecturers and admin. You also need a hospital (and GP surgeries etc) to be attached as well as enough senior staff to train them when they are there. That training is very intensive on trainer:trainee ratios and the senior staff are also in critically short supply as the ones who were trained up when there was training capacity (which, to be fair, was a time when it was far cheaper to train a medical student), are retiring by the thousands and many newer junior doctors quit or emigrate as a result of their experiences up to that point.
Hospitals cost absurdly large amounts of money, especially in the UK where there are consultancies and layers of subcontracting for everything. So the infrastructure costs of adding even a few hundred student places is astronomical.
Due to strategic underinvestment (or ideological sabotage, or governance incompetence, depending on outlook) there is now a self-reinforcing problem: not enough hospitals and staff to train doctors to beef up existing hospital staffing or work in hypothetical new hospitals even if a money firehose was turned on.
Rather than having a tall glass of concrete and doing the hard thing, which will still take decades to manifest, what the government is currently doing is a rerun of Healthcare Assistants where more care is delegated to much cheaper-to-train staff. Plus quite a bit of noise about "AI" bring used to allow them to sweat the asset of the staff they do have by, for example having one radiologist verifying AI findings rather than a pair-based system.
Which will all work "kinda ok" and let them punt the problems at least into the next government's domain when they lose the next election and can spend 5 years screaming from the Opposition benches about the mess. But you cannot do it forever (or the hail-Mary works, there's an AI revolution and you can actually run a hospital with an app, 2 agency nurses, a few smart plugs and an AWS instance).
Of course "doing the hard thing" would be easy to say if it was just the NHS, but there's the same structural degradation in everything. So you also need to spend billions on education. That's the same general problem - shortage of facilities and not enough existing staff to train new staff and the ones you do train quit. Schools are currently one something like a near-500-year replacement rate (50 per year, 24000 total) and that's without considering population growth or even the hundreds of schools that need rebuilding because they're made of RAAC. Then roads need billions to repair the accumulating damage. The railways need huge investment and staffing. Energy is the same - virtually no supply of domestic nuclear design engineers mean they get absolutely rinsed on even squinting in the direction of a drawing of reactor (though Hinkley Point C cost spiral is currently EDF's problem, not the taxpayer and green energy is actually a rare success story). Defence is similar (e.g. ships being retired because they need the staff elsewhere). At least some water networks are collapsing into a multi-billion hole after private dividend extraction. So that money firehose has a lot of work to do, even if they would turn it on. Which is ideological poison apparently.