I have two general ideas. Steal them. I don't have time to implement.
1) Find innovative ways to reduce production costs. Production is expensive. Unimaginably so. This can be things like:
a) Redesign physical set equipment. Ex: Set lights are expensive; a single light can cost tens of thousands of dollars. Design a new light that takes advantage of modern LEDs or something, or that can be deconstructed to serve multiple lighting functions (beyond what exists today).
b) Design automation to replace physical support crew on the set. Payroll is an enormous cost. Ex: Invent a tracking system that would make the boom mic not require a person to physically move it around to follow the actors. Yes, mechanization would eliminate jobs - such is the nature of creative destruction.
c) Innovation in use of set design and photography to make things look much more expensive than they are. A shining example of this was how Orson Welles shot Citizen Kane. Another example is what Spielberg did with the nonfunctional shark in Jaws. For those two, you're probably going to have to check out a book to learn how they did it. And more recently the impressive-looking but cheap sets and photography in Underworld (2003). These directors did unbelievably good things with constraints. Find a way to systemize this. To study what could be made cheaper, look at films that cost way too much, and see how each shot could've been cut down to bare bones - the absolute minimum dollar amount required to make the exact same look and effect.
The most choking example of waste I've ever personally witnessed was on the set of Van Helsing (2004) in Prague. The entirety of Old Town Square was cleared, about a hundred security held people out around the perimeter, and giant cranes and lights lit a scene of Hugh Jackman leaping in the air. What was the end product that appeared in the movie, you ask? A two-second shot of Jackman leaping in the air above some cobblestone. The whole thing could've been accomplished on a 10' by 10' mockup of cobblestone in a studio, or a cobblestone street somewhere cheaper to rent the entirety of than the Old Town Square of Prague.
2) As mentioned elsewhere in this thread, marketing costs are gargantuan. You have to get eyeballs on your enjoyable entertainment.
I watch independent film. The difficulty of going into a little-known film, even one that's received some good notice, is that I generally don't know a thing about it.
People enjoy trailers. Create trailers. Once production is complete, making a trailer for it is essentially free - it requires the software you used to edit the film and your own time and effort.
Run strings of trailers for unknown films on IFC. Run a 30-minute block of 30 to 45 second teasers with links to the film's site, where you have the 2 1/2 minute trailer and more information. Get eyeballs on productions. Right now getting eyeballs on productions is forbiddingly expensive. Revolutionize it. Cheapen it.
I don't think innovating other entertainment will kill Hollywood any faster. You have to innovate the film industry itself. Every human culture throughout history has enjoyed watching performers enact a story, of some kind or another. Moving pictures of performers enact a story isn't going anywhere in popularity. There's a reason it's such a profitable industry. Technology has advanced to the state it's in today where moving pictures of performers enacting stories are easily viewed the world over, but I don't think you should think about changing the type of entertainment so much as the mechanics of the medium.
On the set lights/LED issue - designing a new light that uses LEDs would probably INCREASE the cost, not decrease it, since LED lights are currently more expensive to produce than the alternatives.
That said, if instead you had a company that provided the rental of these lights at a lower cost that normally paid to purchase them outright, perhaps you could offer lower cost lighting with modern, less-breakable technology.
Of course, LED lights are a different type of light from incandescent, etc. anyway, so it's possible that they'd be totally unusable anyway.
I interpreted his LED idea to mean that a lighting system could use multiple types of LEDs, providing video data in specific frequency bands. Software filtering could be used to isolate these, then recombine them in different proportions to get the right 'feel' in color.
If different types of LEDs were positioned far enough apart, you could possibly achieve different lighting angles from one scene-shot. Compare this with shooting a scene multiple times while having the crew reposition lighting. Such a system would indeed be more costly, but could reduce human repetition.
I read halfway through point number 1 and I thought "robotize the studio." Think of a new studio that is infinitely configurable with lights and cameras, and all controlled by software. Kind of a Holodeck but in reverse, or a blackbox theatre where the point is to record everything that goes on in the studio and configure that recording, maybe even something similar to a 3D scanner but with full motion cameras. 3D Models and location overlays could be inserted more easily because all the details of the original shots are already recorded, and can be adjusted after shooting.
What this is would be a standardized studio room, that would only require the director and actors on set. Every other specialty could be done in post, and thus allow them to work on more projects at once. Better skilled people, doing more work, drives quality up and lower costs. Win:Win
I can cut 40% of a film's budget just going outside of the studio system. Plus more with technology you reference.
Studios want the films to appear to lose money so they can keep all the profit. They control distribution and inflate production costs >35%, move money from one pocket to another etc all to limit the net profit participation of equity holders.