Support NeoGAF

[FoxMcChief]

Yeah, tens or hundreds if thousands of devs and creatives are going to learn to weld or do other trends.

As far as welding, you know that is being automated as well. So I guess welders need to do what, construction?

Everything is getting automated, which again proves my point that the game industry is just finally dealing with it. Either adapt or move on. That’s life. Deal with it. Construction is also starting to see automation. Have you seen the 3D printed homes made of cement? Very cool stuff.

At the end of the day, people should worry about needs more than wants. My current job isn’t my career, but it’s stable and pays extremely well and not likely to be automated. People need to let go of the “American dream” concept.

 

[mdkirby]

It is what it is. Maybe they should try to branch out and learn to weld.

They might not even lose there jobs, they could be spun out into another team as publishers are suddenly able to take more risks again....or if there are big layoffs, that's a lot of talent sitting around now able to form new studios to make games with far smaller budgets than what was required previously.

Don't get me wrong tho, there will def be some layoffs, or requirement for retraining, as some roles at big studios full of niche specialists will just become obsolete and require different people.

 

[ByWatterson]

more people lays off

And a much healthier industry. AAA simply isn't sustainable without cost-cuts, and the best way to do that while maintaining complexity and quality will be AI.

I hate to break it to everyone, but AI + designers is objectively good for the industry and the consumer.

 

[Tams]

He's not even hiding it. I mean, just look at him; he's corporate Robbie Rotten.

 

[Wildebeest]

AI is a broad field, but when people hear about it, maybe they imagine something like an exec firing a game director and typing in ChatGPT "come up with a AAA game idea for me and write the script".

 

[diffusionx]

If AI becomes a trend in videogame development, it's not going to "speed up" development, any more than the idea of hiring 500 contractors in Bangalore sped up development. It's just going to let devs churn out stuff faster, which is going to just mean that they can do more, which means the standard will be raised, which means that if anything the price and time will continue to go up. Because if ShitDev#484 can use AI to make their game look better and more polished, so can Sony. So can Rockstar.

 

[Hudo]

Cool. The first tool they should replace with AI is the CEO.

 

[Clear]

Pretty amazing this guy still has a job, does anyone, anywhere think EA have grown in industry stature since 2013?

 

[Men_in_Boxes]

I'm starting to think CEOs can't win here...ever.

NeoGAF: "We want shorter dev cycles. AAA budgets are too high."
CEO: "OK. AI can help us with that.:
NeoGAF: "Not like that!"

Lots of mud slinging. Very few solutions.

 

[diffusionx]

I'm starting to think CEOs can't win here...ever.

NeoGAF: "We want shorter dev cycles. AAA budgets are too high."
CEO: "OK. AI can help us with that.:
NeoGAF: "Not like that!"

Lots of mud slinging. Very few solutions.

AI won't make shorter dev cycles.

Skull & Bones took 10 years to make. Starfield took eight years to make. Dragon Age Deadname has been in development since 2015 according to Wikipedia and maybe will come out this year. Yet somehow Sony made 2 Spiderman games in four years, and also a massive sequel to God of War that was like 3x the size of the first one (or at least feels that way) and got it out four years later. Now four years is a long time, but to me the clear issue here is management.

 

[Deerock71]

More AI involvement will lead to LITERAL developer hunger.

 

[simpatico]

I'm starting to think CEOs can't win here...ever.

NeoGAF: "We want shorter dev cycles. AAA budgets are too high."
CEO: "OK. AI can help us with that.:
NeoGAF: "Not like that!"

Lots of mud slinging. Very few solutions.

There’s really a battered spouse syndrome going on between certain gamers an AAA pubs. You try an impress upon them the indie uprising and they spit it right out. “No, I want my character action slop in a very specific way. Not like that. I need state of the art puddles, ambient occlusion and beloved pop culture characters”. When the UbiSlop Slop Wars game releases, it will be very interesting to cross reference that OT with these AI corpo bashing threads. Because let me tell ya, there is nothing on earth more corpo than a UbiSoft Star Wars game. My prediction is most people on the anti corpo side are going to be day 1 buyers.

 

[N1tr0sOx1d3]

 

[Men_in_Boxes]

AI won't make shorter dev cycles.

Skull & Bones took 10 years to make. Starfield took eight years to make. Dragon Age Deadname has been in development since 2015 according to Wikipedia and maybe will come out this year. Yet somehow Sony made 2 Spiderman games in four years, and also a massive sequel to God of War that was like 3x the size of the first one (or at least feels that way) and got it out four years later. Now four years is a long time, but to me the clear issue here is management.

I'll take cherry picking for 200 Alex.

 

[Mithos]

I think what he means is that there is a hunger among CEOs to cut development cost, so they can increase their own bonus.

I wish there was someone ballsy enough to say this to their faces during a Q&A like this...

 

[hybrid_birth]

Gives their name Electronic Arts a new meaning.

If AI is going to create games.

 

[ReyBrujo]

If AI becomes a trend in videogame development, it's not going to "speed up" development, any more than the idea of hiring 500 contractors in Bangalore sped up development. It's just going to let devs churn out stuff faster, which is going to just mean that they can do more, which means the standard will be raised, which means that if anything the price and time will continue to go up. Because if ShitDev#484 can use AI to make their game look better and more polished, so can Sony. So can Rockstar.

I guess you aren't a developer or you aren't using ChatGPT or Copilot to code if so. The difference between 500 contractors in Bangalore and Copilot is that AI turnover is much, much quicker to obtain and review. For those who don't program the usual development cycle when having contractors is that you tell them to do something, they estimate a time like 2 hours or 4 days after which they upload a PR (or pull request) to a branch in the repository, that PR is reviewed by an in-house developer and either is accepted and merged or rejected with a comment to fix either a possible bug or style. That contractor might be working for three different companies each with a different style guide so he might be mixing them (like using all uppercase names for constants instead of using CamelCase, or returning by reference instead of returning tuples when returning multiple values, etc. They also might create helper methods that already exist but they have never used before leading to code repetition, some of which might not even be caught by the in-house reviewer (which adds the possibility of bugs because you modify one method thinking it's the one being used everywhere but it turns out there was an almost exact copy being used somewhere else).

Copilot parses the whole trunk and understands immediately the style guide. So when you request something it outputs code that "looks" OK in a matter of seconds. It's your task to review and check, but it minimizes many instances of mistakes. For example, it knows there are already helper methods that it can call instead of building its owns, it can write and modify testing code so that you can upload the PR with tests covering the modifications, if you find something that's looks sketchy you can request a change and it will "fix" it immediately.

Of course, all these systems are still experimental and not perfect so the reviewer needs to be careful when accepting code (which is why I used some double quotes around) but it's not different from reviewing code written in Bangalore except that the changes you request are done immediately while you are "in the flow".

What does this mean for the software industry? Well, for starters junior positions will be scaled down, and it will be harder to enter the industry because many of the tasks you brought juniors in for are now done immediately and with a better quality by these tools (something mentioned in a video by Nick Chapsas, wish I could find the link again). What about those who are already in the industry? Well, there are jobs that will be cut undoubtedly, the software industry has a lot of redundancy and if there's something programmers hate is order and timetables (which is why programmers don't like project managers or scrum masters or whoever is taking the managing position). Is anyone safe at all? Of course, you cannot ask ChatGPT "write me a Red Dead Redemption 3 game" just as you cannot ask it to build you a bank or healthcare system. If you ask me the ones that will stick around are those who adopt these tools and exploit them to the maximum. I don't know anyone who is still punching cards at work or writing code in ed one line at a time, nor people coding without an IDE (I, and guess many, sometimes do because it's faster to edit a single file to modify a single line using vim than launching Visual Studio 2022, but it's not the norm).

Does this mean we will be getting two FIFA or two NBA or two annual games per year now because development times are faster? Not really, it doesn't make sense for companies to release faster iterations of the same game just because your developers are now 50% more productive. However it will mean that games will have more time of testing (which is an area that has always been considered an afterthought until books like Software Testing Techniques by Boris Beizer and The Art of Software Testing by Myers in the late 70s/early 80s were published) and that there shouldn't be buggy releases like Cyberpunk or similar.

Again, I think gamers are being more Catholics than the Pope, finding a problem where developers themselves find a solution. In the end, why should gamers care whether the game is using Unity or Unreal, if the game is coded in C# or C++, if it has a code coverage of 35% or 80%, etc. 70% of developers approve of AI usage, why gamers think it's bad for the industry?

 

[Griffon]

AI is a way to gain time. Not just for EA or Activision, but also for small studio and indies. Devs shouldn't shy away from it and use it to its full potential.

 

[Dr. Wilkinson]

...Yet not a hunger among gamers, interesting.

 

[NickFire]

I'm starting to think CEOs can't win here...ever.

NeoGAF: "We want shorter dev cycles. AAA budgets are too high."
CEO: "OK. AI can help us with that.:
NeoGAF: "Not like that!"

Lots of mud slinging. Very few solutions.

If you're keeping tabs put me in the column of people who have no objection to shortened dev times because of AI.

But I am not a developer. If I were a developer I would think any colleagues who shared my (non-developer) view were either paid off or stupid. Just the tip never stops at the tip. The AI that makes a dev's job easier today will probably replace them tomorrow.

 

[Randall365]

It wouldn’t surprise me if AI could knock out the next version of Fifa - name changes, new title screen, new music… DONE

 

[StereoVsn]

What does this mean for the software industry? Well, for starters junior positions will be scaled down, and it will be harder to enter the industry because many of the tasks you brought juniors in for are now done immediately and with a better quality by these tools (something mentioned in a video by Nick Chapsas, wish I could find the link again). What about those who are already in the industry? Well, there are jobs that will be cut undoubtedly, the software industry has a lot of redundancy and if there's something programmers hate is order and timetables (which is why programmers don't like project managers or scrum masters or whoever is taking the managing position). Is anyone safe at all? Of course, you cannot ask ChatGPT "write me a Red Dead Redemption 3 game" just as you cannot ask it to build you a bank or healthcare system. If you ask me the ones that will stick around are those who adopt these tools and exploit them to the maximum. I don't know anyone who is still punching cards at work or writing code in ed one line at a time, nor people coding without an IDE (I, and guess many, sometimes do because it's faster to edit a single file to modify a single line using vim than launching Visual Studio 2022, but it's not the norm).

Again, I think gamers are being more Catholics than the Pope, finding a problem where developers themselves find a solution. In the end, why should gamers care whether the game is using Unity or Unreal, if the game is coded in C# or C++, if it has a code coverage of 35% or 80%, etc. 70% of developers approve of AI usage, why gamers think it's bad for the industry?

My issue is that I see job losses. The junior problem you highlighted is already happening.

As AI tools get more productive, more precise, developers will start getting canned. Yeah, who doesn’t like co-pilot, ChatGPT (4 Turbo preferably) or other similar tools from say Anthropic? They make coding and scripting (and writing documentation) a lot smoother.

I personally don’t write anything complex (mainly Python, terraform, AWS and azure specific scripting, powershell) as I am more on architecture side, but I work closely with many devs and I support bunch of cloud infrastructure for AI project utilization amongst other crap.

And already as a result of improvements in LLM accuracy over last couple years, hiring got slowed down in comms and junior technical positions.

If you don’t think this is going to make it up the stack you are kidding yourself. Senior Devs are probably safe for quite a while. But mid level folks? Don’t know about that in a few years.

 

[diffusionx]

I guess you aren't a developer or you aren't using ChatGPT or Copilot to code if so. The difference between 500 contractors in Bangalore and Copilot is that AI turnover is much, much quicker to obtain and review. For those who don't program the usual development cycle when having contractors is that you tell them to do something, they estimate a time like 2 hours or 4 days after which they upload a PR (or pull request) to a branch in the repository, that PR is reviewed by an in-house developer and either is accepted and merged or rejected with a comment to fix either a possible bug or style. That contractor might be working for three different companies each with a different style guide so he might be mixing them (like using all uppercase names for constants instead of using CamelCase, or returning by reference instead of returning tuples when returning multiple values, etc. They also might create helper methods that already exist but they have never used before leading to code repetition, some of which might not even be caught by the in-house reviewer (which adds the possibility of bugs because you modify one method thinking it's the one being used everywhere but it turns out there was an almost exact copy being used somewhere else).

Copilot parses the whole trunk and understands immediately the style guide. So when you request something it outputs code that "looks" OK in a matter of seconds. It's your task to review and check, but it minimizes many instances of mistakes. For example, it knows there are already helper methods that it can call instead of building its owns, it can write and modify testing code so that you can upload the PR with tests covering the modifications, if you find something that's looks sketchy you can request a change and it will "fix" it immediately.

Of course, all these systems are still experimental and not perfect so the reviewer needs to be careful when accepting code (which is why I used some double quotes around) but it's not different from reviewing code written in Bangalore except that the changes you request are done immediately while you are "in the flow".

What does this mean for the software industry? Well, for starters junior positions will be scaled down, and it will be harder to enter the industry because many of the tasks you brought juniors in for are now done immediately and with a better quality by these tools (something mentioned in a video by Nick Chapsas, wish I could find the link again). What about those who are already in the industry? Well, there are jobs that will be cut undoubtedly, the software industry has a lot of redundancy and if there's something programmers hate is order and timetables (which is why programmers don't like project managers or scrum masters or whoever is taking the managing position). Is anyone safe at all? Of course, you cannot ask ChatGPT "write me a Red Dead Redemption 3 game" just as you cannot ask it to build you a bank or healthcare system. If you ask me the ones that will stick around are those who adopt these tools and exploit them to the maximum. I don't know anyone who is still punching cards at work or writing code in ed one line at a time, nor people coding without an IDE (I, and guess many, sometimes do because it's faster to edit a single file to modify a single line using vim than launching Visual Studio 2022, but it's not the norm).

Does this mean we will be getting two FIFA or two NBA or two annual games per year now because development times are faster? Not really, it doesn't make sense for companies to release faster iterations of the same game just because your developers are now 50% more productive. However it will mean that games will have more time of testing (which is an area that has always been considered an afterthought until books like Software Testing Techniques by Boris Beizer and The Art of Software Testing by Myers in the late 70s/early 80s were published) and that there shouldn't be buggy releases like Cyberpunk or similar.

Again, I think gamers are being more Catholics than the Pope, finding a problem where developers themselves find a solution. In the end, why should gamers care whether the game is using Unity or Unreal, if the game is coded in C# or C++, if it has a code coverage of 35% or 80%, etc. 70% of developers approve of AI usage, why gamers think it's bad for the industry?

I didn't say that using AI is comparable to hiring contractors, I said that it's not going to "speed up development" when everyone can do it, it's just going to raise the baseline of what is possible which companies will take advantage of. Like, if I can hire 10 Indian contractors for the price of 1 American, even if you got 10x the amount of work done (which you don't), well that just means everyone you are competing with can get 10x the amount of work done too, which means that if you want to stay afloat in the industry you need to hire 10, which means you really need to hire 20, which means the costs continue to climb.

When everyone can speed up development, no one can speed up development. I mean, to use your example, the fact that we have computers that can do a gazillion more operations a second than the punch card ones did not mean that less people went into the field or made employees irrelevant because computers can punch cards on a transistor instead. Yes we have less people that need to punch cards (right now: zero), but it opened up the opportunities for more work and to the point where anyone can do far more.

Another instructive example is I remember when baseball really started getting into analytics, people said that scouts were done, it's over, computers could crunch all the numbers and pluck out the good players with just data. Well, not only did scouts not go away, teams have a way bigger staff doing player personnel and analysis than they did before analytics. And they also spend way, way, WAY more on it than they used to. And the initial premise of analytics - moneyball, cheaper teams can erode the advantage that richer teams had - turned out to not be the case as the richer teams could use analytics too, and spend more on it, and spend more on the players the analytics discovered.

If every company had a "click this button and generate Red Dead Redemption 3" button, then everybody can make RDR3, which means that RDR3 is actually the *least* of what customers will accept, which means that companies then need to invest heavily to top RDR3, which means that the competent people and teams that can top RDR3 become that much more valuable, which means there is still a heavy incentive for people to become good at their jobs and make true art over AIslop.

 

[FoxMcChief]

Reads more like some people here fear progress. I just wonder if the reception had been different if it came the President of Nintendo.

 

[ReyBrujo]

My issue is that I see job losses.

If you don’t think this is going to make it up the stack you are kidding yourself.

Yes, lots of whip makers lost their jobs when Lincoln abolished slavery. Also lots of blacksmiths lost their jobs when Henry Ford introduced the Model T. Yes, lots of women lost their jobs as computer operators (the ones that punched the cards) because computers started using magnetic tape (which in time created the belief that women were not interested in computing in general and programming in particular along that Cannon and Perry's paper). Yes, lots of toll booth operators lost their jobs when automatic ones were introduced. Trains and subway drivers are next, eventually programmers.

The solution is not stopping progress, the solution is training those at risk of losing their jobs due obsolescence because someone, somewhere will use those tools to their advantage and your company will eventually catch it up with a massive shakeup or will just fade into bankruptcy (or to be absorbed by a bigger company that will eventually shut it down). What about those wanting to join the (software) industry? Well, even though the bar is pretty high nowadays the area is still expanding. Back when I started in 2002 having a degree or studying for one was enough to be hired, right now they are expected to code in 2 or 3 languages if full-stack, or to have knowledge about the designing process.

I have been in the healthcare industry for a long time so I'm probably more valuable due my knowledge about the business knowledge than my programming abilities. There are lots of laws that a software needs to comply to be legal, there are a lot of standards to follow to get approval by the different agencies, there are a lot of services that need to be consumed and offered for interoperability as requested by the different governments, etc. And knowing all of that gives me an advantage even when compared to someone who might have more experience in other industries or more experience designing or programming. The same can be said about software for banks, or software for many other industries.

Video game programmers don't accumulate that business knowledge? If a video game programmer works for EA for 20 years don't they accumulate some kind of knowledge that puts them above anyone else? Why would a parent company close a successful branch? I obviously sympathize with those who were let go but wonder if they were let go not because of a problem with the people but with the profession. Maybe the job of a game developer is too generic, is too bland and can be replaced by any other developer regardless of previous experience?

Like, if I can hire 10 Indian contractors for the price of 1 American, even if you got 10x the amount of work done (which you don't), well that just means everyone you are competing with can get 10x the amount of work done too, which means that if you want to stay afloat in the industry you need to hire 10, which means you really need to hire 20, which means the costs continue to climb.

My boss would ask me how long would something take and I would answer, 2 weeks. He would ask me if I could get 3 others to help me and have it down in a couple of days and I would say that just because a pregnant woman delivers a baby in 9 months doesn't mean 9 pregnant women will deliver a baby in one month. Software is not just writing code but integrating it with the existing code base (which might lead to integration hell) and that's the problem when you develop too much too fast, you add several layers without having finished testing the previous ones and when something fails you will not which change introduced the bug and have to revert back.

If every company had a "click this button and generate Red Dead Redemption 3" button, then everybody can make RDR3, which means that RDR3 is actually the *least* of what customers will accept, which means that companies then need to invest heavily to top RDR3, which means that the competent people and teams that can top RDR3 become that much more valuable, which means there is still a heavy incentive for people to become good at their jobs and make true art over AIslop.

That's exactly what's happening with China, they develop so many similar games that they end up flooding every genre (for example the action RPG one). They all look generic, almost as if they all had been done using "Action RPG Maker", yet they have a certain quality that would require any small company wanting to join the genre a game that has at least the graphics and the system these cheap copies have. And right now the only way to do it is using AI to leap the advantage those companies already have against yours.

I'm not defending management, mind you, just want to understand why these things happen to game developers. You don't usually see banks wiping out a complete branch when they released a good product.

 

[Tams]

If AI becomes a trend in videogame development, it's not going to "speed up" development, any more than the idea of hiring 500 contractors in Bangalore sped up development. It's just going to let devs churn out stuff faster, which is going to just mean that they can do more, which means the standard will be raised, which means that if anything the price and time will continue to go up. Because if ShitDev#484 can use AI to make their game look better and more polished, so can Sony. So can Rockstar.

The difference is that shit devs could only make shit games up until now.

But now shir devs will be able to make poor games. Which almost certainly means we're going to get a plethora of poor, half-arsed games, but they'll still sell and thus the industry will lean that way.

 

[Fafalada]

He's not even hiding it. I mean, just look at him; he's corporate Robbie Rotten.

According to the Stack Overflow survey of last year 44% of developers already use AI tools, 26% were going to start soon and 30% were not planning. And 77% of developers had favorable feelings regarding AI. So, my conclusion is that the people here complaining about AI replacing jobs actually aren't developers, or have no idea about what AI actually is but fear it will replace their own jobs

I use AI tools at work regularly myself - as you note, they can help productivity in many ways when used right.
But you have to have your head stuck really deep in the sand to not see the writing on the wall for labor market at large now. What will actually come out on the other side noone knows - but we're well past the point of no return now.

 

[Hawking Radiation]

If AI can supplement the development of AAA games without losing the workforce then I'm all for it. Giving developers tools to increase productivity and shorten development cycles is a huge PLUS for me.

Not sure why this would be viewed as bad.

 

[Bernoulli]

AAA games in 3-6 months soon

 

[diffusionx]

The difference is that shit devs could only make shit games up until now.

But now shir devs will be able to make poor games. Which almost certainly means we're going to get a plethora of poor, half-arsed games, but they'll still sell and thus the industry will lean that way.

A shit dev will still make a shit game, because the definition of shit is a moving target.

If one company in a vacuum was adopting this stuff they'd have a huge competitive advantage, obviously, but if every company is doing it, then nobody has a competitive advantage insofar as the tools exist. The primary difference is how the people at the company use it, key word, PEOPLE. So in other words it's the same dynamic as it is now, it's the same dynamic as it was when desktop PCs came out, it's the same dynamic as every transformative technology.

 

[Bernoulli]

AI is a broad field, but when people hear about it, maybe they imagine something like an exec firing a game director and typing in ChatGPT "come up with a AAA game idea for me and write the script".

But it works like that, soon devs will be able to ask it to make a character instead of drawing it

 

[Silver Wattle]

If developers ever make a union, one of the things in every agreement should be executives can never speak for them.

 

[ReyBrujo]

the writing on the wall for labor market at large now

Indeed, a full-stack is one guy doing the job of a front and a back-end developer, a devops is someone doing support, infrastructure, development and even coordination. This ("consolidation") has been happening since quite a long time by now, it didn't start with AI. Companies that cut their workforce in half or two-thirds and expect the ones left being able to deliver the same output as their previous teams are in for a rude awakening.

But now shir devs will be able to make poor games.

The mobile phone environment didn't need AI to be flooded with cheap knock-offs And considering a huge number of Switch releases are just straight copies from their PC/Mobile phone siblings I would say it doesn't change the status quo: lots of companies delivering half-assed games and a few ones releasing good ones.

If developers ever make a union, one of the things in every agreement should be executives can never speak for them.

Interestingly we have three in Argentina and developers don't want to join them. They consider them archaic. Since the software industry is still expanding you can leave a job for a better one and you can obtain better deals when being contracted. Salaries are above average and the unions cannot match the raises software developers can get by themselves. This is specific to Argentina, though, but consider that down here unions are huge, much more powerful than the ones at US in terms of political power.

 

[MrRenegade]

This guy reminds me of G-Man from HL. rofl

 

[sachos]

it will never be as creative as a human being.

You can’t actually create anything unique if you don’t feel anything.

Ehh im going to disagree with the majority feeling in this thread. The creativity is in the prompt. And it is not only about art, as long as coding ability and context length keeps increasing it will be a massive help for programmers and bug fixing. Even if artists refuse to use AI to create final art for the game they can still fine tune internal models with their art style to have a quick way to brainstorm creative ideas, take the best concepts and start manual drawing and refining based on that.

Look, i know people will lose their jobs, its inevitable... but i have hope that AI can trully help bring dev cycles down. And i hope those devs that lose their jobs can use those very same tools to create much better indie games than otherwise could.

Now things could get weird if Sora like models keep improving to the point you can run it in realtime at 30fps based on user input, at that point everyone have infinite playable worlds in their hands. But even then i still think people will like dedicated classic games.

I do think we will need some kind of UBI or something though, otherwise noone will have money to buy those games. Or maybe money isnt a concept anymore in a post scarcity society. Interesting times ahead.

 

[IAmRei]

As a dev (albeit not in gaming) assisting tools are a step forward in the right direction. Back in the 50s and 60s programmers coded in paper, then gave the printed pages of code to an operator who would use a typewriter-like machine that punched holes in a card to then feed a computer each of them to run the program. In the early 70s developers used one-line editors until someone realized they could use the whole screen to edit files and not just one. Until the 80s people coded using text editors until IDEs appeared, allowing them to compile, run and especially debug code. Then came intellisense and now code generation. All those tools are there to make the developer focus on the core, the business logic, the thing that actually makes a product valuable. There is a huge amount of boilerplate code (sometimes necessary, sometimes not really) that is extremely boring to deal with, and with boring things comes disdain and carelessness which opens the door for bugs.

According to the Stack Overflow survey of last year 44% of developers already use AI tools, 26% were going to start soon and 30% were not planning. And 77% of developers had favorable feelings regarding AI. So, my conclusion is that the people here complaining about AI replacing jobs actually aren't developers, or have no idea about what AI actually is but fear it will replace their own jobs, or are being more Catholic than the Pope.

Mostly if you talking stack, its programmer, system engineer, and logic based work.

As for artist, it will be huge stab for them. And they are also needed for video games to be worked. you only see game = programmer in your statement above. Its not valid, half valid.

And game without art is like product without soul.

I'm game developer in audio visual and game design, mostly art parts, but i also handle in core team as game designer and creative direction. I'm not new in this industry either, i works since NDS era. And i own small game developer studio which i cant tell who and what.

 

[diffusionx]

Ehh im going to disagree with the majority feeling in this thread. The creativity is in the prompt. And it is not only about art, as long as coding ability and context length keeps increasing it will be a massive help for programmers and bug fixing. Even if artists refuse to use AI to create final art for the game they can still fine tune internal models with their art style to have a quick way to brainstorm creative ideas, take the best concepts and start manual drawing and refining based on that.

Look, i know people will lose their jobs, its inevitable... but i have hope that AI can trully help bring dev cycles down. And i hope those devs that lose their jobs can use those very same tools to create much better indie games than otherwise could.

Now things could get weird if Sora like models keep improving to the point you can run it in realtime at 30fps based on user input, at that point everyone have infinite playable worlds in their hands. But even then i still think people will like dedicated classic games.

I do think we will need some kind of UBI or something though, otherwise noone will have money to buy those games. Or maybe money isnt a concept anymore in a post scarcity society. Interesting times ahead.

AI has no creativity because it cannot create. All it can do is output something based on the algorithms somebody coded into it and the data somebody fed into it. This is just fundamental. AI can't produce anything new. It doesn't have a mind or a soul, it's just a machine, and can only do what it is told to do.

 

[DunDunDunpachi]

 

[CuteFaceJay]

By the looks of it all the developers are getting sacked off so the CEO's can tell the AI to make the product.

I have a bad feeling that ”AI” in gaming will be used to push out non-stop massive piles of trash games that all look the same, akin to fast food because no matter how sophisticated AI can be, it will never be as creative as a human being.

You can’t actually create anything unique if you don’t feel anything.

Much like the crash of trash in the 80's

 

[daclynk]

 

[T-0800]

AI please update roster and up the number on the cover by 1.

 

[RoboFu]

I have a bad feeling that ”AI” in gaming will be used to push out non-stop massive piles of trash games that all look the same, akin to fast food because no matter how sophisticated AI can be, it will never be as creative as a human being.

You can’t actually create anything unique if you don’t feel anything.

That's already been happening now with " indie " titles that aren't really indie games but some clone treadmill developers pushing out super cheap super shitty games who's only selling point is there name kind of sounds like a popular game and their logo kind looks like a popular games logo.

 

[Porcile]

Yes, I'm sure devs can't wait to produce games more quickly so that can get fired or shut down in two years instead of five.

 

[FewRope]

The guillotine has a "real hunger" for clueless CEOs

 

[StereoVsn]

Ehh im going to disagree with the majority feeling in this thread. The creativity is in the prompt. And it is not only about art, as long as coding ability and context length keeps increasing it will be a massive help for programmers and bug fixing. Even if artists refuse to use AI to create final art for the game they can still fine tune internal models with their art style to have a quick way to brainstorm creative ideas, take the best concepts and start manual drawing and refining based on that.

Look, i know people will lose their jobs, its inevitable... but i have hope that AI can trully help bring dev cycles down. And i hope those devs that lose their jobs can use those very same tools to create much better indie games than otherwise could.

Now things could get weird if Sora like models keep improving to the point you can run it in realtime at 30fps based on user input, at that point everyone have infinite playable worlds in their hands. But even then i still think people will like dedicated classic games.

I do think we will need some kind of UBI or something though, otherwise noone will have money to buy those games. Or maybe money isnt a concept anymore in a post scarcity society. Interesting times ahead.

Yeah, all the hundreds of thousands (or millions in the wider industry) devs are going to create indie games or other independent software and people will buy all of them… sure.

We already have too much content being produced between games, TV, books, and other forms of entertainment so very few flows to the top and are successful. For games alone, there are thousands being released on Steam. Some good and great games don’t make it and sell in very low numbers.

Now, multiply that by AI produced/assisted development. Like you said who the hellos going to buy all that if people lose their jobs?

Oh, and having UBI or another alternative gov sourced income that will provide a decent life? At least in US I don’t see a shred of possibility for that.

 

[FunkMiller]

I'm starting to think CEOs can't win here...ever.

NeoGAF: "We want shorter dev cycles. AAA budgets are too high."
CEO: "OK. AI can help us with that.:
NeoGAF: "Not like that!"

Lots of mud slinging. Very few solutions.

"I won't take a bonus, and I'll reduce my salary (and the salary of all executives) so that we can have more game dev teams working on more games, from different genres and gameplay styles, at different budgets, in a timely manner, that doesn't require them to crunch. It may take some time for the results to filter through, but this is the right decision to maintain this company's reputation and success moving forward."

But, you know, by all means keep defending Andrew Wilson, his 20 million dollar a year pay packet, and his personal 200 million dollar fortune.

 

[Polelock]

I'll be my right nut this is more from stakeholders and executives looking at how to flip assets for quick profit.

 

[Men_in_Boxes]

"I won't take a bonus, and I'll reduce my salary (and the salary of all executives) so that we can have more game dev teams working on more games, from different genres and gameplay styles, at different budgets, in a timely manner, that doesn't require them to crunch. It may take some time for the results to filter through, but this is the right decision to maintain this company's reputation and success moving forward."

But, you know, by all means keep defending Andrew Wilson, his 20 million dollar a year pay packet, and his personal 200 million dollar fortune.

If Andrew Wilson's "pay packet" was the problem, EA wouldn't have become one of the largest, most successful 3rd party publishers on planet earth.

But, you know, by all means keep lusting after another mans salary. Think of how many baseball players the Los Angelos Dodgers could have employed if Shoehie Ohtani halted his 700 million dollar pay packet.

 

[FeralEcho]

Oh look, another cancerous CEO only doing what's good for him and his executive mates.

Hell is gonna be overpopulated with so many of these greedy cunts in the industry.

 

[FunkMiller]

If Andrew Wilson's "pay packet" was the problem, EA wouldn't have become one of the largest, most successful 3rd party publishers on planet earth.

But, you know, by all means keep lusting after another mans salary. Think of how many baseball players the Los Angelos Dodgers could have employed if Shoehie Ohtani halted his 700 million dollar pay packet.

Lol. You think this about personal jealousy?

Wilson's pay packet absolutely is part of the problem, when you take into consideration the number of people who actually MAKE the games who have lost their jobs recently.

Your comparison to the baseball guy is wrong. Ohtani is more like a Kojima or a Miyazaki.