TV is supposed to be ad-funded, and for once it seems the math actually works out.
Using Game of Thrones here for convenience, an episode costs $10-15 million, at a time they were getting 10-15 million viewers. Shows are typically 44 minutes of runtime per hours, which after various dithering gives you about 15 minutes of ads, or 30 spots.
$1 per viewer, 30 spots, need 3.5 cents per viewer per spot for thin profit margins. This is in fact a reasonable price for a spot, albeit on the high end. It's plausibly profitable, though there's a reason TV execs cancel shows at the drop of a hat.
Don't forget the financial point of a TV firm isn't to make money for the studios, but to make money for all the people drawing salaries from the studios. The studios themselves are largely prestige projects.
That said, clearly TV sucks because ads don't work. Ads don't move sales and thus can't pay enough. The studios have to cut every cost possible, which makes TV live up to the high Soviet quality standards all communist schemes live up to.
P.S. I rather expect ads are also prestige products. Ads make you infamous rather than famous, but for some that's better than obscurity. Secondly it signals real-business-ism. If someone hasn't seen your creation on TV, then they're going to think it's a low-class or weirdo purchase. Knowing it exists is a non-problem or isn't solved by ads.
Youtube has a similar cost schedule, "YouTube ads have an average cost-per-view of $0.010 - $0.030" but that money is split between Youtube and the videographer, they don't show nearly as many ads (especially to folk like me), and if you skip the ad it doesn't pay. Apparently "There are now non-skippable bumper ads of six seconds, charged on a CPM basis, i.e., per thousand views." I don't know if you've noticed, but 6 is 1/5 of 30. You can make good money if you can regularly make million-view videos quickly and by yourself ($20,000 per), but in practice million-view channels are softcore prostitution, involve a team, have significant material costs, etc.
I just saw a video that probably made $150,000. Neat. Only it took 40 days, significant materiel, and a team which I know includes at least three members, though I don't know their exact split. For this guy it clearly beats digging ditches, but there's a term I think about a lot: risk-adjusted capital cost. Certainly if you already have a good channel, like, keep going. However, if you're thinking of getting one, don't. The risk-adjusted profit on this venture is almost certainly negative. The opportunity cost is downright obscene. Which makes sense: Youtube is almost always a hobby, and the fact it's a hobby (that happens to pay money) is priced into the compensation. Sort of like this: if you're doing a thing anyway, you might as well try filming the thing.
I figure Youtube is one of the many businesses exploiting the difference between real interest rates and the interest rates being charged, meaning it's a wealth sink. Or pork trough: since money is flowing through, you can scam some off the top. They may also not be including depreciation. It's hard to price depreciation accurately and thus easy to cook the books without breaking the law, sometimes even by accident. Either way if the economy wasn't being buggered, Youtube would immediately go under.
P.P.S. Is Youtube bandwidth is more expensive than TV bandwidth because it's bespoke, or is it cheaper due to lack of redundancy? They don't broadcast to anyone who doesn't request the broadcast. On the other hand, an hour of video marginally costs Youtube something like 5 cents per viewer, in other words a noticeable fraction of ad revenue. Much more like 10 cents per paying viewer.