Antropic ai

ilexev

ilexev

Refugee
Hi everyone, hope this subject wont receive a bad welcom lol, only for reflexion and thinking.
Is it time to use Antropic AI?
Maybe one of the best tools for devs, learn, and be updated in terms of innovation.
Build up and level up the game faster.
Allow to devs to build better work setup and learn faster, having help by technologie, the essence of technologie.
And having better programmeur, developer and ingénieur programmation.
Maybe gain Time for ray tracing, having a good in game store or something like that faster? Futur dlcs ? Paid Dlcs after 4.0? A lots of mini dlc After the 4.0? A mod shop program in partnership with modders?
 
Considering the amount of hate towards AI these days, that wouldn't be a good move. People have already complained about AI use in certain art that was done for this game. Any serious use of AI for game development is likely to continue to receive hate and bad reviews and lack of sales for at least another 5-10 years. Eventually, people will see that good games can be made with the help of AI, but until there are enough out there that are good, it's going to hurt sales and reputation. It will likely be small indie devs with very small games who can say it's one person or some small team making the game and they can't do it without AI, as well as big companies that don't care about their reputation that much since it's already really bad (Electronic Arts) who will be the ones pushing into the use of AI and lead to people coming to accept it. But that won't happen overnight and we are certainly not there yet.
 
Ty, whe know.
AI is used by hundreds of millions of companies worldwide, across various industries. Ethical AI practices depend on its use, such as not stealing the work of artistic creators. But what about the technical, time-consuming, and optimization work involved? What does the general public think? Nothing, because they haven't yet considered the issue. Ethical professional use of AI is possible: no AI voices, no AI-generated graphics, no AI-generated music. But what about the extremely time-consuming, tedious, and complex work? I think it's a good thing for everyone, regardless of public opinion, if it brings more benefits than harm. Regarding pollution, why should average companies in the video game industry suffer more harm than others? Public opinion on this subject is hypocritical.
Post automatically merged:

And complexe and chronophage works* real dev can Always rewatch the work done by ia or programs
 
It really doesn't matter if AI is used in regular businesses. The hate for AI is much higher for anything that is artistic - art, writing, games - or for anything that replaces jobs - robots to work in factories and such. It's far less for most other industries. You may think it's not a big deal to use AI. I'm not opposed to AI myself, though I understand just how limited it really is and how many mistakes it can make, especially in programming, and I know how much extra effort is required to find and fix the problems AI creates in programming that can often make it faster not to use AI for programming. So even though I'm not opposed to using AI, I still don't see it as being a viable option for most developers for at least 5-10 years. As I said, a one person developer might need to use it for things they can't do, such as if they aren't an artist or sound designer and can't afford one, and a company like EA is already looking at replacing their regular developers with AI in upcoming games and has already received a lot of pushback from players. But most devs aren't going to go that route before AI is far more accepted, and that won't happen soon.

Anyhow, only TFP can say what they'll do with AI. They've obviously used it in small amounts (art, at the least) and may continue to do so. But I doubt they'll do any significant work using AI anytime soon. But we'll see.

And you did ask for opinions and discussion. I responded and told you what I thought. You've seen a couple people who didn't want to respond but already marked it as not liking the idea. That's pretty much what you're going to see from players for any suggestion to use AI. Until that changes, it's just not a good idea for most developers to consider for anything other than really minor things.

Regarding why people don't like AI, here are a couple of the main reasons that I've seen....
  • It takes away jobs/work from people. If someone with no skills can do something in a fraction of the time as someone who has skills, that upsets a lot of people. You can find a lot of books written using AI that can be written in a couple of days, while actual writers might need months, for example. Many people think it's not right to take away work from people who put in the effort to learn their skills.
  • Social media is a big cause of hate towards AI. Everywhere you look, you're being bombarded with obvious AI images and videos and posts that are poorly made, obviously fake, and intentionally misleading. People are getting really sick of it, and if that trend continues, it might be even longer before AI use is accepted enough to not hurt a developer's reputation.
In time, opinions on AI will change. But it will definitely take time.
 
Last edited:
I will say that fake movie trailers are annoying and they used to be immediately recognizable since they could only be made by piecing together existing footage. Now, since AI has advanced, there have been a couple times that I was completely fooled and only realized it was fake when I went to IMDB to get more information. This is truly one of the more evil uses of AI. "Thou shalt not rick roll people with fake movie trailers" would have been commandment 11 if there'd been more room on the tablet....or if the Israelites had had movies.
 
Riamus said:
It really doesn't matter if AI is used in regular businesses. The hate for AI is much higher for anything that is artistic - art, writing, games - or for anything that replaces jobs - robots to work in factories and such. It's far less for most other industries. You may think it's not a big deal to use AI. I'm not opposed to AI myself, though I understand just how limited it really is and how many mistakes it can make, especially in programming, and I know how much extra effort is required to find and fix the problems AI creates in programming that can often make it faster not to use AI for programming. So even though I'm not opposed to using AI, I still don't see it as being a viable option for most developers for at least 5-10 years. As I said, a one person developer might need to use it for things they can't do, such as if they aren't an artist or sound designer and can't afford one, and a company like EA is already looking at replacing their regular developers with AI in upcoming games and has already received a lot of pushback from players. But most devs aren't going to go that route before AI is far more accepted, and that won't happen soon.

Anyhow, only TFP can say what they'll do with AI. They've obviously used it in small amounts (art, at the least) and may continue to do so. But I doubt they'll do any significant work using AI anytime soon. But we'll see.

And you did ask for opinions and discussion. I responded and told you what I thought. You've seen a couple people who didn't want to respond but already marked it as not liking the idea. That's pretty much what you're going to see from players for any suggestion to use AI. Until that changes, it's just not a good idea for most developers to consider for anything other than really minor things.

Regarding why people don't like AI, here are a couple of the main reasons that I've seen....
  • It takes away jobs/work from people. If someone with no skills can do something in a fraction of the time as someone who has skills, that upsets a lot of people. You can find a lot of books written using AI that can be written in a couple of days, while actual writers might need months, for example. Many people think it's not right to take away work from people who put in the effort to learn their skills.
  • Social media is a big cause of hate towards AI. Everywhere you look, you're being bombarded with obvious AI images and videos and posts that are poorly made, obviously fake, and intentionally misleading. People are getting really sick of it, and if that trend continues, it might be even longer before AI use is accepted enough to not hurt a developer's reputation.
In time, opinions on AI will change. But it will definitely take time.
Click to expand...
I think you haven't independently explored AI and haven't formed your own opinion about the AI market and what it offers. Anthropic AI was released this week, and its capabilities have blown away all benchmarks. It's a developer-focused tool that far surpasses others, which is why this thread is named after it. This AI, released this week, could prove truly useful for many developers and programmers. Of course, don't touch art and graphics; handmade items are imbued with the soul of their creator, as are books. The performance of this AI is so high that developers won't need to double-check every line of code, etc. Furthermore, this AI integrates new, functional execution tools that you should test to understand what it's really about. It's not counterfeit. There are people who are not ready for dozens of big, obscure and absurd reasons, and normal people who question themselves at least twice a week.
Maybe Perhaps a transparent charter or regulation on the ethics of AI use? A controlled use, authorized only for technical, time-consuming, and complex tasks.
 
Well, since you asked, using AI as a programming tool is inevitable in the software industry. It will destroy jobs and it does in a way steal from creators and it will generate other problems, but it is inevitable because it is more efficient. So whatever my opinion about AI, there is no use arguing against it. And the advantages, especially in programming, are obvious, as it is the solution to the generation of the much hated "boilerplate code". I used it for that as well. But it is not a miracle cure and it also takes time to adjust to it and neither will it necessarily increase the quality of the code.

TFP has programmers that already know about AI, like everyone in the industry. If they want to use it they have done that already.

And my opinion about in-game shops and paid mods you already know.
 
I'd be surprised if TFP wasn't using AI to some limited extent, Anthropic AI being a good choice.

That said, I would suspect the usage would be limited because I'll bet TFP has a pretty large code base and I suspect AI will have trouble understanding how it all fits together, conceptually, and compartmentalizing subsystems.

When I try to involve using AI with the interpretation of compliance standards I have to be mindful that it can't tell what it read in each version of the standard and will sometimes give an out-of-date answer. I suspect it might have that problem with Unity. You would want to keep an eye on it as it might want to start using a library that TFP wants to avoid. It might give you expedient answers that aren't reentrant.

All that said, AI is an easy answer if you want a quick Python program to make some kind of repetitive change to POI XML files. I use an AI-generated Python script to generate Localization.txt for every language except English for my POIs. You could do that for an analysis of world files.

AI is pretty handy at explaining unfamiliar code and that can be a nice time saver.

As for how the profession of programming and software engineering changes, well place your bets. I think we'll see some of those roles transition into less-well trained "citizen" developers, and that the hiring of junior programmers will rebound to some extent. If you don't hire junior developers you will eventually have no senior developers. Again, my opinion.

One of the things we've been messing with in classes here is the use of the Model Context Protocol and writing MCP Services for Claude (Anthropic AI) to use. I'll explain that...

Folks are used to conversing with an AI in a web browser. As a web page, the AI is contained within the browser's sandbox, so it cannot manipulate your computer. But if you install a native application (claude-code for example) then the AI runs with whatever level of user privileges you have when you launch the application. That AI can come with "tools" that let it do things. For instance, on Linux it comes with a "bash" tool, which lets it issue bash commands. If you write an MCP Service (not a lot of code for a basic one) then you can give the AI more tools.

For example, I'm certain Claude's training has informed it of how to run a 7D2D server and the details of SteamCMD. I'd bet if you had claude-code installed on a Linux box that you could ask to do things for you and it would use the bash tool to do them.

Looking at possibilities, if you were to add an API to 7D2D or to some tool such as Teragon such that a Python program could invoke that API and make things happen -- like doing something in game, or configure Teragon's options -- then you could write an MCP that gave Claude the ability to invoke that tool. Imagine having a conversation with Claude, like...

"Please create a 4k 7D2D world with the default ground height of 60 that is mostly forest. I want these POI mods enabled: A, B, C. No traders."

"Connect to the game, go into Admin mode, teleport to where JoeGamer is standing, and issue the command to kill all the nearby zombies."
 
zztong said:
I'd be surprised if TFP wasn't using AI to some limited extent, Anthropic AI being a good choice.

That said, I would suspect the usage would be limited because I'll bet TFP has a pretty large code base and I suspect AI will have trouble understanding how it all fits together, conceptually, and compartmentalizing subsystems.

When I try to involve using AI with the interpretation of compliance standards I have to be mindful that it can't tell what it read in each version of the standard and will sometimes give an out-of-date answer. I suspect it might have that problem with Unity. You would want to keep an eye on it as it might want to start using a library that TFP wants to avoid. It might give you expedient answers that aren't reentrant.

All that said, AI is an easy answer if you want a quick Python program to make some kind of repetitive change to POI XML files. I use an AI-generated Python script to generate Localization.txt for every language except English for my POIs. You could do that for an analysis of world files.

AI is pretty handy at explaining unfamiliar code and that can be a nice time saver.

As for how the profession of programming and software engineering changes, well place your bets. I think we'll see some of those roles transition into less-well trained "citizen" developers, and that the hiring of junior programmers will rebound to some extent. If you don't hire junior developers you will eventually have no senior developers. Again, my opinion.

One of the things we've been messing with in classes here is the use of the Model Context Protocol and writing MCP Services for Claude (Anthropic AI) to use. I'll explain that...

Folks are used to conversing with an AI in a web browser. As a web page, the AI is contained within the browser's sandbox, so it cannot manipulate your computer. But if you install a native application (claude-code for example) then the AI runs with whatever level of user privileges you have when you launch the application. That AI can come with "tools" that let it do things. For instance, on Linux it comes with a "bash" tool, which lets it issue bash commands. If you write an MCP Service (not a lot of code for a basic one) then you can give the AI more tools.

For example, I'm certain Claude's training has informed it of how to run a 7D2D server and the details of SteamCMD. I'd bet if you had claude-code installed on a Linux box that you could ask to do things for you and it would use the bash tool to do them.

Looking at possibilities, if you were to add an API to 7D2D or to some tool such as Teragon such that a Python program could invoke that API and make things happen -- like doing something in game, or configure Teragon's options -- then you could write an MCP that gave Claude the ability to invoke that tool. Imagine having a conversation with Claude, like...

"Please create a 4k 7D2D world with the default ground height of 60 that is mostly forest. I want these POI mods enabled: A, B, C. No traders."

"Connect to the game, go into Admin mode, teleport to where JoeGamer is standing, and issue the command to kill all the nearby zombies."
Click to expand...
Not forgetting that Claude has new tools that seem quite useful, such as the "tools for tools"and optimizes tokens better than other AIs
 
My I\O interpretation

Ethics and morality of the programmer.
Become the program.
Ethics and morality of those that attain access to the AI.
Become the program.

Ability to not be controverted and compromised once attained by others. Impossible
It can be cloned, even if hard written to a Rom. Examples: Arduino and game bots,
necessity for continuous security updates to an OS and financial establishments,
Social media to attain artificed fame for $. Lower education standards plus more
people being nurtured by the the artificial power and annonymity of typed words,
breeds a skewed offshoot of ethics and morality. Games and media forms, have begun
to cater to that at a faster pace, because it is a lucrative model.

Phishing, Coldcalls, advertisements, and virii are direct examples of controverted
AI programming.

Limited use and access "is only limited in an arena of one", on a non connected device.

But the constant/underlying justification, and often glossed over, issue is ethics and morality.

Oppenheimer. SimEarth. Nielsen Ratings.

Unspoken purpose becomes the end result, it to is inevitable................π
 
ilexev said:
I think you haven't independently explored AI and haven't formed your own opinion about the AI market and what it offers. Anthropic AI was released this week, and its capabilities have blown away all benchmarks. It's a developer-focused tool that far surpasses others, which is why this thread is named after it. This AI, released this week, could prove truly useful for many developers and programmers. Of course, don't touch art and graphics; handmade items are imbued with the soul of their creator, as are books. The performance of this AI is so high that developers won't need to double-check every line of code, etc. Furthermore, this AI integrates new, functional execution tools that you should test to understand what it's really about. It's not counterfeit. There are people who are not ready for dozens of big, obscure and absurd reasons, and normal people who question themselves at least twice a week.
Maybe Perhaps a transparent charter or regulation on the ethics of AI use? A controlled use, authorized only for technical, time-consuming, and complex tasks.
Click to expand...
You don't think I know about AI because I disagree with you? That's not how knowledge works. I know quite a lot about AI and have used it myself, including minimally with programming. It creates a lot of problems because it isn't human. It just takes information from a variety of sources and tries to use that information to do something. In many instances, something ends up being wrong. And you have to spend time trying to find out what is wrong and fix it. Just like it can take a lot of extra work and effort to debug someone else's code, it takes a lot of extra effort to debug AI code for the same reason. If you didn't write it, you don't know how it was written and have to go through it far more carefully to find problems than if it's your own code and you already have a good idea what you did wrong. AI is fine on smaller codebases, but it isn't great with larger projects. It has its uses, and it will certainly get better over time, but it's not there yet. And it doesn't change how people perceive the use of AI.

As has been mentioned, TFP has used AI for at least some things. We know that. We also know there was backlash about it even though it was something minor. TFP will likely continue to use AI to some extent, but they aren't going to just switch to using AI for major programming tasks. At least, it's not very likely. As I said, give it 5-10 years and things will likely change. But it won't happen today.
 
Roland said:
I will say that fake movie trailers are annoying and they used to be immediately recognizable since they could only be made by piecing together existing footage. Now, since AI has advanced, there have been a couple times that I was completely fooled and only realized it was fake when I went to IMDB to get more information. This is truly one of the more evil uses of AI. "Thou shalt not rick roll people with fake movie trailers" would have been commandment 11 if there'd been more room on the tablet....or if the Israelites had had movies.
Click to expand...
I do the same; I always check on IMDB to see if a certain movie is actually being made, or if someone created a fake trailer.
One other way to know if it's fake is the channel name itself; unless it's a trailer from a major studio, it's usually fake.
I don't know how many times I've seen an "Alita: Battle Angel" sequel coming soon... Not true.
 
Riamus said:
You don't think I know about AI because I disagree with you? That's not how knowledge works. I know quite a lot about AI and have used it myself, including minimally with programming. It creates a lot of problems because it isn't human. It just takes information from a variety of sources and tries to use that information to do something. In many instances, something ends up being wrong. And you have to spend time trying to find out what is wrong and fix it. Just like it can take a lot of extra work and effort to debug someone else's code, it takes a lot of extra effort to debug AI code for the same reason. If you didn't write it, you don't know how it was written and have to go through it far more carefully to find problems than if it's your own code and you already have a good idea what you did wrong. AI is fine on smaller codebases, but it isn't great with larger projects. It has its uses, and it will certainly get better over time, but it's not there yet. And it doesn't change how people perceive the use of AI.

As has been mentioned, TFP has used AI for at least some things. We know that. We also know there was backlash about it even though it was something minor. TFP will likely continue to use AI to some extent, but they aren't going to just switch to using AI for major programming tasks. At least, it's not very likely. As I said, give it 5-10 years and things will likely change. But it won't happen today.
Click to expand...
I stand by what I said; you know nothing about AI. Having a more informed opinion than average doesn't make you an expert. And your personal use and experience don't define anything and aren't representative of the generality
 
I'm not sure why I'm even bothering responding to you at this point, but I'll try one last time...

If you think I don't know what I'm talking about, then tell me what I said that was incorrect? Not based on what you want to believe, but on actual facts. Have you personally tried programming with AI? Has it been able to write large programs or sections of programs without errors? Have you tried debugging a large section of code generated by AI? Was is as easy as debugging your own code? You seem to be saying you're an expert that knows far more than anyone else about it, so have you personally done these things? Because I have.
 
You must log in or register to reply here.
Back
Top