Wouldn't attribute that to HD graphics alone. Game developers have been pushing the envelope of graphics since gaming started. I rather attribute it to many different factors, of which complexity in games (graphics being _a_ factor of said complexity) is perhaps a major component. Making last gen graphics games are not going to instantly make games profitable.

I think companies are at more of a risk than core games.

There are specific quotes from developers highlighting the difference in costs of making HD graphics, including Polyphony Digital's comparison of how many days/months an artist takes to make a car in GT5 as compared to the previous GT games. Graphics is certainly the biggest factor here.

and someone likely considering lair a AAA game at the time

Since Lair is listed at "initial" not the actual price of lair.

wii is destroying the games as we know then, everyone will go zibraz or babyz
or well, all i can say is

"Wii Qube and Wii Relax revealed" :(

Developers will simply have to learn they have to make everything within a reasonable budget.

Sadly they wont learn this until more jobs are lost or a few more companys go under.

That kind of thing is simply a case of using a level of HD graphics that does not fit.  That is a horribly inefficient method for converting a car into a HD 3D model (there shouldn't even BE artists who work on making a car in GT5).  I'm not saying studios shouldn't re think how and where they alocate funds to what parts of the project, but you CAN most definitely create a complex HD game without those horridly high costs.

Any suggestions on how to go about doing this? Because to the best of my knowledge, the best and the brightest in the industry don't have the slightest idea about how to make things more efficient. I'm not mocking you or anything, mind you; I'm being 100% serious when I say that no one, and I mean NO ONE, seems to know how to bring those costs down. I'm not even asking for an explanation, mind you, because this computer stuff tends to go over my head. I am, however, suggesting that if you really see a way to save developers' millions, you need to start sharing with them, if not for their sake then for your own.

in reference to that graph,

I thought that by this generation there'd of been a program for voice recognition software, that would, say, for example:
in old wrestling games(say WWF Attitude) if you want the announcers to say your  created wrestler's name during a match you'd pick from a pre-selected list of popular names, but with the new software everyword would be pre-programed into the console with smart technology picking up on the pitch and tone programmed into the game.  I saw in  an episode of Law and Order, if I remeber correctly, there's a device that upon hearing someone's voice can capture the tone and go on to let the another person say other things through the device with, what would essentially be, that persons voice (I could've been thinking back in my mind to a wrestling game I was crafting when I was 14 with a smart voice recogition system like that.)

My point is:

Developers, when making games, should focus a lot of time on making new tools and engines, in addition to the games they're working on, so that they, and the development community can be more well off in future projects.

Right off the bat, GT5 for instance prides itself on exactness of their car models.  Every single car in GT5 already has an autocad model by the manufacturer that the studio should have.  Viola.  They already have to pay for the rights to have these cars IN the game, they can pay the extra money (if it would cost extra, which it shouldn't) to get the autocad files of the car from the designer, which I assume would be less than the cost of the hundreds of man hours spent on making the one car from scratch

Stop making pre-rendered driven cut scenes.  Instead of spending the hundreds of thousands of dollars on massive render farms to create cutscenes that are at this point barely better than the regular gameplay graphics, just created the scenes using the in game engine.  Its been done already and its proven just as good now.  Hell I woulda been fine with no pre-rendered scenes back in FFVII days.

Stop reinventing the wheel.  Good studios either create a game engine they can use once for several games or they buy one already on the market.  Unreal engine, square enix's crystal engine comes to mind.  Before you mention Sqaure Enix's money problems, they've suffered from a spat of really bad games lately, something being cost effective wouldn't necessarily solve.

And for the love of god learn to hire better programmers.  I'm not saying such and such studio hires crapy developers or ragging on the entire industry.  But over half the programmers and developers of every field are bad at what they do, inefficient at their job at best.

These are some off the top of my head.  Some studios i'm sure already do some of these, some will start.  Eventually the studios left standing will figure it out how to get it done.