Erm, if you take "content creation" to mean full, playable games then you're right. But...
I'm pretty sure Valve is referring to content at every level, from hats to in-game items to custom levels to mods.
His point stands, however, and you only have to look at the quality of many of the Steam Workshop addons produced by hobbyists. Yes, this includes plenty of high poly models and much more, easily comparable to anything produced professionally.
But what I was saying is that these guys are making single models of very high quality, not filling out an entire GTA V5. And especially not coding, acting, mocapping, producing, marketing etc a game. They're not even building somthing like bastion with a small team. They are many orders of magnitude between doing one model that sells well and building a full game. Especially one that does well.
Gabe has mentioned that there are customers that are making $500,000 a year via content production in the Steam Workshop. From his perspective (or anyone's, really), that's simply too compelling to not be a part of the narrative being discussed with regard to having full control (again, both technological and financial, in my opinion).
Clearly, they are seeing compelling experiences (and equally compelling cash) being delivered via user generated content, and it is in their best interest to foster and mature that space.
The source that I read (the quote, below, from Gabe) indicates people, which is of course plural. Also, it indicates $500,000 per year, not over three years:
"So now we're in this strange world where we have people who are using the Steam Workshop who are making $500,000 per year building items for other customers."
Exclusive interview: Valve's Gabe Newell on Steam Box, biometrics, and the future of gaming:
http://www.theverge.com/2013/1/8/3852144/gabe-newell-intervi...
By way of example, it looks like Everquest Next is implementing a more advanced version of Minecraft's building-block technology, which would seem to provide the foundation for allowing players to create whole new worlds in separate servers, like Minecraft but with more sophisticated world building tools.
Any thoughts on whether that might take off?
for that, i would use second life as your measuring stick.
people will pay professional developers for toys to play with (eq next, for example), but enabling hobby developers to make money off their efforts using your platform? that's hard. and expensive. you wouldn't believe the support nightmare much less things like anti-money-laundering laws you have to deal with from country to country. it'll make you cry.
have you ever tried to push a platform update out underneath an already money-making-but-that's-my-livelihood audience? yeah, i wish whomever luck with that.
On the other hand, if you have "fun" tools for users to create content, they WILL create it. Like in Little Big Planet, ModNation Racers and others. It's amazingly hard to make a game, and amazingly harder to make a tool that is fun to use. And with this "fun tool" obviously you strip down all the publishing barriers (technical and legal) that "common game development" have.
as a veteran of the "user-generated" game content industry, i can honestly say 5 years ago, i had that same vision. 5 years of blood, sweat, money and tears later, i can (also honestly) say -- that's not the case. and won't be.
for those of you who've never made (finished and shipped!) a game before, it's an incredibly frustrating experience. you have what would be called an mvp in ycombinator terms inside of a month. "i will crush the rest of this thing out 3 months and ship!" uh... no.
i'm here to tell you that is not the case. the whole "first 90%" vs. the "second 90%" thing is magnified 10-fold in games. it's a really long post for me to explain the myriad of gotchas involved from safely syncing game state between sleep-and-non-sleep-packets-suddenly-flood-in-but-omg-my-clock to btw-you-need-3-times-the-states-in-your-state-machine-oops to why-am-i-getting-3-fps-on-a-faster-device to all-of-the-above-times-multiplayer-exponentially. i plan on writing that post someday, but not now.
anyway.
the prototype flies off your finger tips and then the next 2 and a half years (please see the common thread of 3 years to make a game, from the movie indie gamer [highly recommended!] to that recent atomic brawl post) working through obscure bugs and corner cases because games have more complex interactions than non-games and gamers don't have a clue how to follow the damn happy path when they play.
so, yes. i said 1 month of fun, fast iterative prototype and 2 years of awful, in-the-weeds debugging where your strength of will to finish is tested daily. (insert rovio making 51 games before angry birds here) finishing a good game is goddamnhard. finishing a shitty game is a snap -- just ship your prototype!
what we found was that anyone who has the strength and perseverance to actually build and (more importantly) finish a game was already working in the game industry -- mostly because they possess a proven will to finish.
i guess what's i'm saying is lots of people can snap together sweet, use-with-care prototypes ugc-style. almost nobody -- and i really, really mean almost nobody, like 0.0001% of "i want to make a game" developers -- can actually finish and launch a game regardless of scope.
mind you -- i don't view indie games as ugc. if you ship (SHIP, not start) an indie game, you have what it takes to work in the non-indie game world. you just choose not to.
if you don't believe me, i challenge you to prove me wrong and just make a simple game this weekend and let your friends play it where you can watch, but not talk or help them at all. it will take approximately 12 seconds of them touching your game before you're completely "wtf -- how did that happen?!?!?" screwed.
m3mnoch.