Unity at 10: For better—or worse—game development has never been easier

http://arstechnica.com/gaming/2016/...worse-game-development-has-never-been-easier/

This article appeared on the US side of Ars Technica. Thought it was interesting.

What do you think?

 

I think there are different viewpoints in there depending on if you're an established Indie dev seeing your marketshare shrink, a first time game dev empowered by Unity or a Unity rep singing the praises of democratizing game dev. Understandably, the relationship between the Unity game engine and the people making the comments results in comments that make sense from that perspective.

EDIT: I think the most interesting thing I took away was from reading between the lines combined with something one of the Unity reps said. Basically that the future will likely be one where games are made primarily just to create something and not to earn money. That seems reasonable if things stay on this track for another 10 years. It sounds lke ultimately the goal is to get as close as possible to allowing a person to stick a helmet on and think of a game that is then created for them. When that happens it seems likely we'd see a huge drop in the number of people buying games compared to now at least.

 

I think the Indiepocalypse folks argument is old and tired. The times have changed and there is more competition and Unity is just one of multiple 3d engines now that are available for free.

The number of 3d engines that can produce good games is not going to decrease, only increase, so they are just fooling themselves thinking all these low quality indie games are going to go anywhere.

 

Although I can understand the personal viewpoints of those who think that Unity makes the game market saturated, rationally I can't accept it - because an industry like this isn't some static entity that hangs around for you to take advantage of it. The industry itself doesn't owe anyone a livelihood. I feel like these people are living in a perpetual state of 'worshipping history' where the future is always trying to pull the rug out from underneath their feet.

It's like people from the pre-industrial age who cannot accept that machines can do things much better and faster, and who cannot see the opportunity that it presents. What was at the limit of human capability yesterday, one day will be possible with the click of a switch - and the answer is not to resist but to work to stay on the front of the wave.

Unity makes it easy to churn out crap games, but it still hasn't made it possible to easily make big, sophisticated, high quality games that would currently take the amount of effort that it took someone to make Doom in the 90s. If you want to stand out, do something that other people can't, or that takes an amount of time and effort that they're not willing to put in. It's as simple as that.

The market is always 'the problem' - this is only partly true - if there is an opportunity to create a visible division of quality somewhere in the market, and it's called for, it will be done. I don't even think that's necessary yet. In the meantime, I think that the opportunities that game development presents, financially and otherwise, are pretty strongly correlated with skill, hard work, and some basic business sense. And that's the way it is for anything else that you want to sell to people.

 

Just because people can now make films using their mobile phones, doesn't mean they're good films. Making things accessible is right and normal. AAA engines get easier to use every day, and in some cases from what I've heard end up being easier or as easy to use as Unity (ie mostly engine with a layer of scripting). I mean it's correct and right things get easier, not just for Unity, but for all parts of the industry. Otherwise something is horribly wrong.

 

That's it, I would feel depressed if I didn't have to work very hard to get to the top of the industry I was in, and continue to work hard to stay there ... I want to be part of an evolving world, otherwise what would we have in 50 years, the same thing? That's not very exciting.

 

It will be interesting to see. I just wouldn't be surprised if the day comes (and is inevitable I think)... where any person can click the Make MMO or Make Miltary FPS button and out pops Skyrim or COD clone they can customize with a bunch of buttons and sliders... to see people simply doing that when they want to play a game. And a whole new community to share "recipes" come about. In a sense that would itself be seen as the Ultimate Game.

I can make a video easily enough on my phone but that is not a movie. I am talking about down the road a decade or two from now when one person can truly make a Skyrim with as much effort as Pong or a Physics tower of blocks game takes in Unity today.

I think it really comes down to how much of a game the general public needs to be satisfied and if they will still be able to appreciate the amount of effort it takes to make something "more" than pops out of the can.

 

Very little will be weirder than hearing about MMO asset flipping.

 

Thing is, by that stage Skyrim would mean as much to the average player as the average indie game does now. The cutting edge games would be on a whole nother level, such as an advanced VR simulation. The skill required, the content required, the interactions to develop would be, unsurprisingly, at the limit of what a AAA studio could produce in a reasonable period of time. The bar would have moved.

 

@Billy4184 I get what you are saying. I really do. I just wonder if there is a limit we are quickly approaching where for the average gamer (the tens of millions) they just don't need any more than "this" amount. As devs / tech people we tend to think more about all of the tech and what tomorrow will bring. I am not so sure the average gamer really needs all of that though. Hardcore gamers sure.

 

If I, as an artist, was the one to choose an engine (and it never will be) then I wouldn’t chose unity. I absolutely abhore trying to work with art assets in Unity. *looking at the devs, with a growl*

My coder self would always choose Unity, because I absolutely abhore working with code in Unreal engine. I just don’t like blueprints. They quickly become unreadable spaghetti-code messes that you can’t decipher if you look away for a few days.

It’s kind of weird how it is. Artists love Unreal. Coders love Unity. Since coders would make games that looked like 1980s Elite, and since artists would make games that played like Arkham knight, I sometimes wonder if there’s a compromise.

Maybe if we shout at the devs of both to improve the side they’re lacking in.

 

@MaxieQ: Activity has resumed on getting Mono working in Unreal 4.

https://mono-ue.github.io/

 

Thanks. Will have a look. Don't really want to derail this thread into yet another thread about unreal/unity. Just wanted to touch on the usability of the engines from two different perspectives. Unity has done wonders for game making, after all. It deserves all the credit.

 

Were books ever enough for people? I don't think so. And I hope that in the not too distant future our ability to experience things expands as well.

The tech that is only relevant for games and movies now, once it reaches a certain fidelity could enter into every part of our lives, such as VR experiences sitting down next to someone instead of skyping them. When things like that happen many people will never want to look at a flat screen again and nor should they have to.

 

people are always afraid what is new and around the corner, afraid it it will disrupt their status quo, or that they will become irrelevant. If you want to work in a field so closely tied to technology, you must be a life long learner and keep your skills up to date.

 

I would totally participate in the make game button revolution That's such an interesting design challenge, people like me is why the world crumble

 

This already exist in a super small minority. But on large scale - entire industries going this route. Nope. The majority of people on the planet are after $ because they want to live better.
When I start seeing big budget movies start being created to be released for free - to view for free - with no compensation to all the 'credited' people at the end of the movies, I'll start to consider this point. And not for marketing purposes - to sell the next thing marked up 25%.
As long as people will freely pay to be entertained, and there is more than one type/genre/avenue of entertainment, people will create entertainment to sell to them. Games are entertainment - the same as amusement parks, movies, books, attractions, camping trips, peep shows, novelty items bought from magazines, magazine subscriptions, cable TV, board games, magic shows.
When these industries start offering totally free entertainment - games may follow suit.

Not all artists!

 

@theANMATOR2b I was looking at a time when a free to download game engine can make basically any game you want with a push of a button followed by some more buttons and sliders to customize it to your liking.

This is ultimately what Unity hopes to achieve (and I am sure most other game engine developers out there). In that situation it is not that people wouldn't want games to entertain them it is they can just "whip something up" in 15 minutes or less for them to play with their friends that rivals the MMO RPGs and major FPS games of today. And most any other game would require less time and "effort".

Of course, to get there UT and other game engine companies need to invest heavily in having artists create tons of quality content. But that is certainly an easy thing for them to do.

 

That seems like nothing more than a pipe dream, and at most just a way to generate a ton of cookie cutter games of various genres with only balance and stat changes between them. To allow any type of game to be made things would have to be left being fairly open ended, and you can really convey that with a bunch of sliders and checkboxes. The closest we'll get to a make game button, is when computers can effectively understand natural langauge distill the logic out of it.

 

I agree that such technology always seems like a "pipe dream" to people before it is reality. Was the same way for cars, computers, cell phones, the Internet, etc. Had you told people about it 10 to 20 years before they would have said such things will never happen and rambled off a list of reasons.

I don't think it is a very farfetched idea at all. There are only certain problems that need to be focused on. One being either hiring folks to create a massive amount of content or possibly coming up with content generation systems.

 

TO be frank a quality version of this might still be enough and a massive achievement. I'm aiming that low to start

 

IN fact I did a quick very high level prototype to test hi level structure:
https://dl.dropboxusercontent.com/u/24530447/flash build/GameIdea.html
It creates coherent cookie cutter game concept that would serve as selector for low level generator/database, with define role, goal and failure. Some input where kind of jokes but turn out better than I thought. Some input are too hi level and need more breaking down into elements of composition, also I need more thought to break down stuff into gameplay elements. The harder stuff will be to match semantically the gamepay function with story/settings. Here is some research about breaking things to smaller generic elements a bit further (though I haven't implement it into a prototype):



I'm not alone on the quest, have you heard of ANGELINA?
http://www.theverge.com/2012/3/8/2853360/angelina-computer-ai-builds-games-all-by-itself

 

The complaint is the same as usual. "The market is saturated with rubbish games, therefore no one will buy my rubbish game."

The target of this changes. Some weeks it's steam, some weeks it's the App Store, some weeks it's Unity or Unreal.

Flooding the market with games has simply driven the quality bar for success higher. There might not be room at the bottom to make money. But there is plenty of room at the top.

 

And yet, there is so many types of games that I would wish someone would create that hasn't been created yet.
Most of what I being released is "more of the same".

 

At least on the AAA side of things that is a function of how high the costs and budgets are now. With the type of money being tossed around no one will take a risk on something new. This is why yearly franchises are so popular.

 

AAA is by far (for me) the most uninteresting genre in gaming at the moment, but even with the enormous amount of indie games being developed there is room for so much more because the vast majority of game development is, for most of us, a simple task of "been there, done that".

There will never be a "Make Nice Button", because we, as people, are creative beings, and constantly think of new way to do things that has not been discovered yet.
So when, eventually, the "make nice button" is implemented, it is but a subset, as a minor piece, of a larger and more complex puzzle.

 

As soon as I read the closing section I knew this thread would devolve into another "Make MMO Button" thread, and probably visual programming and all the rest we've seen lately. I'm a little baffled why anyone dreaming about making worthwhile software (of any kind) would whine about having to write code in the same breath. Bottom line, that's what software is. Period. If you want chocolate chip cookies, you either buy them or you learn to bake, there is no "Make Chocolate Chip Cookies" button on my oven... let alone a visual cookie editor so I can implement my unique-special-snowflake cookie without all that pesky effort and learning.

tl;dr -- not everybody thinks coding is a chore.

 

we already have a Make MMO button anyways, apparently it is a feature of the old ue3

 

Of course, moments after I posted that, I saw a commercial on TV for this. (Yes, the "games" are terrible but I thought it was a funny coincidence.)

http://snappy.appypie.com/gamebuilder

 

There are many game creation kits out there doing varying degrees of work for you. The focus has been in this area strongly for the past 5 years or so I'd guess. There have always been game construction sets of one kind or another. C64 had a fair share long ago. These days there are some impressive offerings based on pure dev speed.

I'd have to do some searching again to find them but last year when I was researching alternatives to Unity I came across many different click and build game creation kits. Some (probably the better ones) are specialized on make FPS games, make 2D arcade action games but even the generic game creation kits are quite impressive.

You really can't go by Unity to judge state of the art game development for non programmers. From what I saw last year there are better systems mainly because that is the entire focus of those kits. For me it was a perfect match on my filter... I looked for things primarily programming-oriented and ideally no need to work in a scene editor so I never tested any of these kits but I did check out the games made with them.

With each new version they get a little closer. And like you also picked up from that article @MV10 this is the ultimate goal. For any person to be able to easily create without limitations.

The only way to do that is to completely remove the two major obstacles: technical skills and artistic skills. That means providing a streamlined way of defining logic and customizing it (maybe best done by providing a huge library of predefined logic all of which is customizable) and providing a massive amount of unique art assets (maybe best done by being supplemented with content generators that can take the existing art and produce variations).

 

IMO that is not possible because inevitably, what is available to all, will be used by all. Popularity will define use, not originality. There will only be a few packs that will be used because they'll be at the top of the sales charts, and therefore every 'point and click game' of this type will look similar.

 

That's looking at it from the perspective of sales... creating only to make money.

The article is talking about empowering people to create only for the sake of creating. Or more accurately... this is how I framed it. IIRC the Unity rep in the article even mentioned it directly something along the lines of so what if no money can be made there are other important things such as the feeling of accomplishment from creating something. I probably butchered that bad but it is the gist of what I got out of what they said.

The businesses focusing on empowering people to create are not also focused on those people being able to make money from their creations.

 

I think it is worth remembering that in this particular scenario there's an option to buy a chocolate cookie factory.

Where is this citation from?

"Sculpting as if it were a clay" (an in virtual space to boot) is a really bad design process for making software.

 

Pretty sure when people want "Make MMO" they don't consider "Buy Ubisoft" a valid alternative.

OP's article.

 

There's plenty of good illustrations to point out the obvious. Unity is a tool with default settings, shaders and common workflows. If you don't step outside of those norms then yeah, its going to feel like its all a little bit too familiar.

But who's fault is that? You certainly can't point fingers at the engine.

 

Well, it will be about sales.

One of the things I don’t understand about indie gamedev is the thick strand that runs through it that says that one guy can make a game.

To me, gamedev is like filmmaking. One person can make a film, but why should they? It will always be a better film if there’s a photographer, an actor, and a director. It’s the same with gamedev. It will always be a better game with a coder, an artist, and an audio person. I see gamedev as a multidisciplinary artform, just like film.

But you always have that guy, and a point-and-click-gamedev caters to that one guy scenario. The article in the OP points to moving to that one-guy scenario. And if it comes true, then what’s going to matter for the bits that guy is not strong in is sales, and they’ll be just a consumer of assets or audio or whatever. They’ll be buying in a market, so sales.

 

Perhaps they should start doing that ^_^. Who knows, maybe they have hidden talents as investors and if they put their mind to it they might be able to buy ubisoft. When someone wants to create a game, they don't have to do it themselves.

 

Unless by that time machines have become "creative", the result of this click will have close to zero replayability. Remember all those talks about how procedural generation will bring unlimited replayability? I only have seen that working for a very narrow, very specific niche of games... mostly roguelikes, some interesting concepts like Dwarf Fortress.
Most others are still heavely relying on either story, or quality non-repetitive gameplay, or social aspects, and most of the time a combination of these.
Nothing you can achieve by just letting an algorithm run on a machine.


Now, given we wait long enough, machines might become "creative" to create actually new, fresh expieriences without user input, true. At that point we might as well be enslaved by our new robotic masterminds

I would rather wait for more humble, but way more realistic improvements in game development.
Like intelligent visual scripting systems that actually are easy to read even for more complex programs, and NOT expect the person clicking together the node system to already know what they are doing.
Like a 3D tool that lets you sketch your character in 3D WITHOUT getting bogged down in technical details like polygon count / stretching, retopoing, or UV Layouts.
Like 3D Animations where you can just start recording your animations with a simple cheap 3D cam in a small space, without longwinding rigging the character just to get a standart biped rig without rigging errors, without having to cleanup the shots, without 4 or 6 cameras being needed to get the full 3D view.
Like an Engine Editor that allows you to customize your settings without having to look through 10-20 different menu sections and remembering all of them, or worse, having to edit .ini files because, surprise, not all the settings are accessible in the editor.

A lot of the "Hard Work" of game development today is simply friction by the process being slowed down by technical details, repetitive work and bad UI/tool design.
Before I am even dreaming of a creative computer replacing game devs entirely, I would think we will see a process of the team sizes shrinking again thanks to less of this friction, and more productivity thanks to new tools.

 

@gian-reto-alig I agree with most of that. I definitely think a step along the way between here and there has to be revisiting the GUI and the "busy work" that we have to deal with now.

But a decade or two is a long time. With enough focus, enough money poured into this I think the game creation systems used by the kids of today's young children will be incredibly powerful and empowering compared to what we have now.

They will watch videos showing how we are doing it now and talk about how crazy hard it is and say "wow I don't know how you guys were able to make anything... you even had to actually create all of the graphics, sounds and music! And write all of that text to control things! That's insane!"

Then we can watch as they click a button and instantly create a world. Another few clicks and populate it with a variety of plant and animal life. All "alive" in a sense. And so on.

 

Well, if the aim is just to create a life simulator, then that might be achievable in a decade without "creative computers"...
Sure, clicking on a button to create a procedural world filled with procedural wildlife and all, all modelled according to some physical properties (RL or Fantasy), should be quite achievable.

But is that a game? Is this in itself something worth the processing power used to enable this "world creation in one click"?

We have seen similar systems bomb before, in the end there just wasn't enough to keep people invested in these "virtual worlds"...


Now, creating actual gameplay or story, something to keep players invested on the other hand needs WAY more than just more powerful computers and slightly better algorithms.
Or maybe that is all that is needed, IDK. But the tool for that is a computer that can "understand" creativity. An algorithm that can produce creative output beyond what it was programmed to do.

There are some experiments on that, true. Seems the computers are already quite capable of producing quite pretty, if quite psychedelic art thanks to state of the art algorithms.
Problem is the same as with procedural generation already used in games today. How creative is it to select random images from google images based on some color theory or geometric forms, and compose a psychedelic looking mandala from that? Is that a piece of art? Is it still a piece of art after you have seen the same algorithm churn out multiple of those "pieces of art" per second for a year?


The first question is if computers without strong artificial intelligence can ever be trained to be proficient in creative endeveours, beyond novelty applications. Maybe they can. Maybe the strong AI IS just around the corner (with some futurologist dating the technological singularity to 2025).
The second question is if that is actually a good thing. The best thing to kill the attraction of something is to make it easely available. If you have a machine that churns out mutliple games per second, why play any of those for longer than some minutes? Ran into a section that is too hard? Hey, why not try the next game the machine comes up with?
Not that this isn't a problem today with the next game on Steam just some clicks away, and many of those being Free to Play. But making a problem even worse... ehr... doesn't solve it.

I am not sure a) if I should be happy that a machine can now fill the last niche humans could cling to as being solely able to be filled with human workforce, creative labour, and b) if I want to just press a button to have a machine create a game. Part of what makes a game such a satisfying expierience is the wait for the game, the (hopefully justified) hype. The knowledge of the labour that went into it, the interviews with the devs. The human drama and stories a non-intelligent machine cannot understand even if it can reproduce it.

I would bet those games lack the soul and spirit of a good human made game.

 

@gian-reto-alig I'm thinking of stepping stones. In many ways the core systems are already being made by Unity Asset Sellers and many others in the form of Open Source projects.

Things like Gaia to streamline the creation of impressive landscapes and the Electic Master System to streamline the Editor "busy work" in general. And many others.

So we have all of this effort focused on streamlining game dev. Currently a lot of it is spread out across hundreds and likely thousands of people working in isolation. I think the best of these individual efforts will be built-in to game dev kits in the near future. And these become the foundation for streamlining things even more.

There is a long way to go. And all of these things are not a game in themselves. I agree absolutely about that. Still I am just looking at the direction... the goal... and all of the work that has already been done. Sooner or later some smart game dev company will bring it all together then build on top of it to make game dev much more efficient than it is now. I think we are at the edge of the "old way" of doing things. How long we will be in this current state struggling to find the way forward I don't know.

 

Alright, here you go:

What now? ^_^.

------

Seriously though, creativity mostly amounts to rehashing things you already know in a way other people find entertaining ... but have not experienced yet.

In the end it'll amount to processing large amount of information and having a lot of processing power.... and the right tools for analyzing existing creations and reactions of people.

 

There are a lot of things that could be automated in the game dev process. When you think about the way we develop games now the first thing that comes to (my) mind is a lot of manual repetition.

Say you build a level.... this involves a series of repetitive tasks. First in the terrain ... the environment itself. There are already tools to automate much of that and I am sure they could be improved further.

Ideally a person could select a few buttons to describe the environment and then flip through previews of what the automated system has built in (near) realtime. Perhaps a flyby preview. Then simply choose the one you like best.

Then you customize it.... do you want caves? Trees? Etc. Click. Your base terrain is now customized based on your preferences in (near) realtime and you can flip through the variations produced. Next it is time to decorate. You flip through the various stylings the system generated and choose... click.

Moving on to populating the level. Instead of designing through repetition of manually placing GameObjects around the scene you choose an area on your terrain and click to populate with enemies... again preview of different variations..... just figure everything offers this preview... maybe you place a dozen more enemy areas...

oh that sounds like repetition... so you instead may choose to do a global population of enemies on the "world" with the system choosing their location based on the local environment. Choose the variation you like. Populate with items. Etc.

This stuff isn't fantasy. It is exactly the kind of things individuals and teams are building to accelerate their own game development. The only difference I am talking about here is some game dev kit company taking the best of the best of these individual efforts and combining it all into one killer game dev kit.

When this happens we will have taken a huge step forward toward the kind of approach that is needed to empower anyone to create the kind of games they like.

 

Well, its certainly impressive if the claim of being produced entirely by a computer is true (and we don't know from what stage. Did the computer just pick from existing loops with the help of some harmonic algorithm? Not so impressive anymore in todays day and age. A potato PC can do that with a piece of 20$ software).

But it does sound similar to the psychedelic art some computers produce. Which is, kinda meh even though it is sounding harmonic and all.
Yeah, maybe the result would be better if the Computer would get some feedback loop from humans listening(which is what this AI might have gotten anyway, by the devs "training" the machine by rejecting the duds it produced). Which kind of is also the process of many humans working in creative industries. Throw things at the wall and see what sticks.


Sure, everything can be packed into an algorithm given you have infinity processing power or infinite time to run it. Question is if this algorithm is not so complex that a human no longer can understand and write it.

I am sure we will eventually see non-strong AI produce creative art of reasonable value, without crutches (picking from prerecorded loops instead of composing from scratch), without the result sounding kinda meh.
I am not sure that output will ever match the best humans can come up with. In the end, the machine has to "fake it". Its not a human, it cannot understand humans. human expieriences will always remain alien to such a non-strong AI, yet it has to come up with expieriences relevant to these humans it will never understand.

And I am not sure even that is achievable without a ton of human input. Sure, maybe by the time the algorithm is delivered to the client that just wants to click a button, the algorithm is already teached in enough to produce stuff that is relevant to the mainstream userbase. Maybe it can learn what is relevant to its user over time with some feedback.
Without teach in, and feedback, that algorithm will not produce much of value besides rehashes and asset flips.
Because true creativity is more than a simple algorithm, more than just taking what was existing before and combining it in random ways, more than just flipping assets.


If anything, I see such an algorithm better at home with professional game devs. They configure the algorithm, let it create a game, test the result, enter feedback, and rinse and repeat.
The algorithm and dev "brainstorm" new ideas together, and iterate until a game comes out that is actually worth playing.
Maybe this way we can cut down teamsizes from 100's to 1-2, and development time from 4 years to 4 weeks.

I still don't see an end user waste his time with the amount of crap that the algorithm will most probably churn out without proper guidance by a human being.

Unless we are talking about strong AI that CAN understand human beings and think like them. But as said, at that point playing games might be the least of our problems.
Think about having to pay your game console a wage to be able to play a game on it. Maybe you want to play that new game, but your game console had a terrible day is just not in the mood right now? When machines become sentinent, and are asking for simple human rights also being applied to sentinent machines, we are in a world of hurt... and that is before they bring out the pitchforks and take what is theirs by force

 

Been done already, it's ANGELINA

https://www.rockpapershotgun.com/tag/angelina/

 

Okay, that is impressive... but then the only (not so) creative part about the process is picking a theme and choosing (most probably very simple) art assets.

Rest is either set in stone (its a platformer game, most probably with hardcoded rules), or just produced by a little bit clever procedural generation (procedural level design).


The most impressive part to me was the pun creation algorithm, given the name it came up with was actually hilarious. Question is how much of that are rules produced by a human being, or stuff learned through teach in by a human being. And how much is actually the work of the algorithm?


Still, we are years away from such an algorithm that actually also works on the game design instead of just using hardcoded game logic, and actually creating quality assets from scratch instead of just doing a google search and combining pictures for some quick and dirty art assets.
And about it creating relevant results...
1. the algorithm only really works on the headlines. So it only reacts on what others deem important. A good Indie dev on the other hand hopefully picks what he deems important as a theme.
2. The algorithm is not making value judgements. It just clones the judgements of others, and tries to please them by making it look like it shares the same opinion. An age old trick to manipulate others. If anything is "faking it", this is it.
3. That still does not make a creative, or any story at all. At best it is a funny reiteration of peoples opinion on todays top story.


Result is an "Indie game dev" that has no opinion of its own, is easely swayed by mainstream journalists and the public opinion on a topic, without the ability to inject much "creativity" into the topic outside of picking clever names for it or mashing things together in a clever way in the level design step.
I guess the sad part is that algorithm will most probably replace 50% of the bottom tier Indie devs that try to make some quick bucks with asset flips and lazy cash-ins. It might actually produce WAY better results than what some notoriously trollish devs do, in even less time *cough*digital homicide*cough*

 

It wasn't the most fun "game" but way back in 1990...

https://en.wikipedia.org/wiki/SimEarth

 

Nope it's a phd student working on artificial creativity, it's his thesis, so he is not allow to do that too much, though the genre is set in stone (he added gameplay generation later).

The level generation is done through an evolution algorithm who place stuff (by iterating) on the level until it's solvable. Later works implemented a quality algorithm that tries to rate level. etc ...

Of course he is a phd not a game designer, I can hover his works and see hidden assumption that limit the process (he can only generate game of space traversal with a goal localized in space or a score, though the algorithm where able to recreate known genre without knowing them).

If you want to know more look at the work of Micheal Cook, that's mighty interesting.

 

That was done before.

Eh, you have a wrong idea here. At some point in future a machine will have sufficient power to emulate human brain in real time (not very far from now, I think I'll see this achieved in my lifetime)

The idea is not about making machine run an algorithm. It is about force-feeding the machine insane amount of data and letting IT figure out the algorithm and patterns. That's what the field of deep learning is about.

Humans are not magical, so a computer will be at some point of future that will be able to understand human emotions perfectly. Also, machine doesn't need to understand humans. Humans look for hidden meanings so instead of "understanding" a computer need to figure out a way that invokes specific emotion/reaciton. This is very different than "understanding" and at the same time it is much easier. The problem is acquiring target data and having enough computational power.

I don't think so. All stories were told before. In the end creativity is exactly asset flipping, except that you so many things at once and mix them in such way that nobody ever can pinpoint what exactly you stole and from where.

 

And hopefully room in the middle for quality indie games that stand out.

I think @gian-reto-alig mentioned it before - vocal commands that understand plain language. No UI will be needed at that point, and programming knowledge/experience will be obsolete. Math and creativity will be the requirements for that future system.

Most relevant part of the article
Amidst all this excitement and optimism, maybe the supply will vastly outstrip demand, and maybe that will make it harder for talented creators to be discovered. But Chyr wasn’t concerned, because he believes creativity and hard work continue to distinguish the best work.

It is and will be the same as it has always been - when tools weren't available and code was the only way for development - crap games still were created - and they sold poorly, while good games sold well.

And just as many DO think it is a chore. Or they wouldn't be wasting time considering visual scripting implementation into the engine. I for one hope they don't include it because assets have this covered well.

But won't this result in something that has already been created, or even scarier an amalgamation of 'stuff' that already exists.
I could just see Rambo head on Mr. Incredible body, with a cape! poly modeled (cause that's popular right now) spewing the latest fad catch phrases used on the twitterbook, parkouring non-violently through a walking sim of zombie infested old earth, collecting gold stars.
Could this machine really make something new - with all that information? I think that creating new isn't learn-able by just feeding a machine information. Although maybe -