Something i find myself wondering, especially the further "down the rabbit hole" i find myself on this learning venture, "Will ai be able to write "clean" code based off of simple input." My one friend laughed and said i sound like someone from the 80s talking about the future of ai. However i feel that he is more antiquated in his perspective, being "Its just not how it works, and won't".
Once computing was analyzing the results of a serious of input written on paper and reading the output via paper. Then developed into basic input pretty much what we see as machine code, and took a very deep understanding to know what was happening. All that was very cryptic and required a better method. So down the road came all of our input languages, becoming less and less cryptic in nature, and much more readable. So the wonderful people at MIT tried to perfect visual scripting, and that just never took off. I understand why, as it takes such a vast knowledge already of programming and takes the speed element out of the process. Though i believe that was the logical step into making it more human, though it skewed the purpose too far.
Using ai to create art is a wild process, and i find myself laughing at the crazy conclusions it comes to, though it truly has progressed significantly in just a few years. If the image isn't quite what you want, you can just add a level of description and try again. So personally, i believe that the same process could easily be used to take basic thought, and create usable code. Possibly needing minimal tuning (if its a completely "left field" idea), just like art is developed by ai.
I am just curious what "the pros" think about this progression. Is it the next logical step? Or is this just some "Jetsons" mentality about the "future".

Will artificial intelligence replace the need for programmers.

OF COURSE!13%
WELL... not in the near future.88%

    AI will replace humanity. Everything as a species. Including programmers.

    It may replace some programming. It may lead to a new generation of programming languages where you can provide all the instructions in a natural language.

    It won't completely replace programming, because many situations require that a program always work correctly.

      REVBENT Is it the next logical step? Or is this just some "Jetsons" mentality about the "future".

      The latter.

      Maybe not in your time, but certainly there will be new dominant species, if human make a mistake and allow it to be sentient.
      The species will replace humans. Humans just like any other prehistoric thing, will die out. and be replace by somethig else.

      But in near time terms, not really. Maybe in 1 generation there will be new methods in computing and data management maybe a way that you cant imagine just yet.

      Programming in it self is a garbage method in manipulating stuff/data. There should be a better way to do that. More natural way to do it, instead of painstakingly trying to confine your thoughts and wants into some narrow margins defined and limited by a language.

      well thankyou everyone for the replies, good to get the noodle going🙂 i wish the wright brothers had a forum to pass around ideas of flight, and what the ideas at the time were. be truly amazing if those ideas, and the process to development had been properly documented. i really find the perspectives of different individual about where AI is heading super interesting. i agree that the language is always progressing to be intuitive, and surely ai will and in some sense already manipulates code accordingly. I believe we are experiencing a leap in evolution that will not happen at such a rate in any other time in the future. The real "big bang" lol. Personally i see this relationship being more so symbiotic in nature. More so like the mitochondria is the "powerhouse of the cell". given the proper environment, organic life is superior in the recycling of material for perpetuation. Pair that amazing time tested adaption, with the blistering fast extra dimensional group thinking of a computer, and i think that "THING" will be what comes next.

        REVBENT honestly the idea of talking computer is kinda boring and outdated. like you have to interact with it in such outdated way that takes too much time in conveying what you want it to do. Sure telling it in natural way is amazing that it can understand your intention from that alone. But imo the future is more like what Chilong Ma is dreaming of achieving. A perfect BMI (not body mass index) - brian machine interface. where you think it and it does it.

        This is also what Gaben was talking about 10 years ago and he said we can do it now, and he is really looking forward to incorporating this into the steam and stuff.

        We got eye tracking with tobii and stuff. There were some devices that you put on head and it can learn your brian patterns and stuff and it can do stuff what you think.

        I think the AI bros should focus on that, try to create,train models for basic human brain wave interactions, instead of garbage like generating videos/images/or googling or gpt gabage, for just wasting time and electricity where people are starving in some countries and dont have access to electricity etc etc.

        Its possible, like Chin Long Ma said - its not that hard.

        REVBENT where's the NO answer? 100% of AI hype is just ignorant people and out of context quotes.

        REVBENT "Will ai be able to write "clean" code based off of simple input."

        no. 99% of people can't write clean code, what makes you think AI will be able to? and chatgpt crap is made from real people's code run through a glorified random number generator and with no shallow understanding of what's going on.

        DaveTheCoder It may replace some programming. It may lead to a new generation of programming languages where you can provide all the instructions in a natural language.

        It won't completely replace programming, because many situations require that a program always work correctly.

        if anything it will lead to a new generation of really bad programmers who can't do anything at all, and everything will start to fail all the time because people can't do bug fixing.

          Jesusemora if anything it will lead to a new generation of really bad programmers who can't do anything at all, and everything will start to fail all the time because people can't do bug fixing.

          I think that already happened before AI, due to computers being used for everything. But it will get worse.

          Jesusemora because people can't do bug fixing.

          IMO, this doesn't just apply to programming.

          Jesusemora where's the NO answer? 100% of AI hype is just ignorant people and out of context quotes.

          This is dumb. Life always find a way, like making electric cars (EWW). I'm sure in 1000 years it will be possible, but there is no way right now AI can program code. I'm still learning to program guns!

            Jesusemora no. 99% of people can't write clean code, what makes you think AI will be able to?

            This is very true.

            Now can AI fix code, Like Grammarly?

            REVBENT Let's suppose ChatGPT 6 is released tomorrow and it can program whatever you tell it to program. The word "program" here means whatever it meant to you in your original post. Now I'd be interested in seeing the prompt, or series of prompts, you'd give to it to produce your dream game. Can you make a small mental experiment and share those prompts here? I'm also curious about what would be your next step once the ai printed out all 100000+ lines of your dream game code.

            I really like the perspectives individuals have about this concept. Especially the different types of people i speak to about this. The cavemen mentality being my favorite in the sense of doom and gloom, to the people who believe that there is a possible transcendence into a virtual life. I was hoping to have some level of input from the godot forum, and this has been absolute gold in my eyes. I wish i could reach more people with this thought and document the responses properly just as a time capsule of thought and perspective leading up to and during the assimilation to AI. I think it will happen much like the y2k threat (if you are old enough to have experienced that), and something that just opened our eyes to the need for better planning with computers. Meanwhile you have the crazies LITERALLY prepping and throwing out electronics preparing for the END. Something so basically human of us, but evolution favored the ones who prepped harder. The niche is then filled to exploit that preparation by a "smarter" creature (penguins and rocks for nests comes to mind), creating a new threat to prepare for. I get the "need" people have to fear new things, but certainly the terminator concept is extreme, but i hope the human perspective doesn't create a barrier to really accept the benefits AI WILL bring.

            @kuligs2 i feel like the talking computer aspect is a different route, more on the level of visual scripting and may come with some of its shortcomings. the integration of tech into our brains is definitely the "next" big step, and yes, happening as we speak. i more so am asking if the computer will be able to write clean or even cleaner code than we can now, and in what time frame. doing so from the input we give it, what ever means that may be. i think the video generating is more so a method of learning how to "think" like we do, and a process to analyze what we are saying. Possibly a "waste" of time, but all our knowledge (arguably) is formed with a similar method thru our days, from the first one.

            @Jesusemora I felt like a no was just a really closed minded thought, not to be offensive with that statement, but id argue anyone who can handle this level of thinking is very aware the shackles of no/not going to happen bring. BUT.... playing devils advocate, as the lack of allowing a response is more closed minded than saying no to something, i apologize for not allowing the vote. I will certainly add the tally for NO mentally. I guess my mentality is more that human error and the lack of shared knowledge limits us to human productivity. So of course humans will write the code incorrectly and need a lot of time to polish their craft. That is where a computer being able to "perfectly" understand the outcomes, would need to bridge the gaps of human interpretation. Once it can understand what we want to say, it could create the fastest path to doing so, kind of like the fist time python code runs (right?), except doing that however many of times with ABC variables for which method really is fastest/concise. Coding is difficult mainly because we are building on ideas from the past, and using them to solve modern problems. That makes our need to prior knowledge immeasurably important. If ai is used to understand the bug and then sort thru a vast directory of how problems were solved before, i feel at least 90% of necessary code could be done by a program and its learned habits. It would only be the really revolutionary ideas that would need some human input to bridge the understanding gap. As far as flying cars, we do have them, but the functionality really isn't there. I had a deep conversation at a wedding with a guest i had met there (jet salesman), and he made a great point that to create something you are safe being hit by another vehicle in, the whole being able to fly part becomes a big factor... just to heavy. i believe there is an english company thats sold them for some time now, but only to the island hopping superrich outside of the masses. The infamous "They" recently made a battery powered flying suit, like iron man kinda... that's more space age to me than a flying car.

            @lukboy i think that the "like grammarly" statement is down the path i'm addressing. Using that same process to how code "should" be done, be that simple error assistance, could certainly be done with enough info properly indexed and "AI'' to analyze it. Kind of like The rust language, The blistering speed of C but with the lack of assigning/releasing memory done by machine (I could be totally speaking out my behind, but i think that's how that goes), just doing that same thing on a much higher level from basic thoughts and a Grammarly style approach to making basic thought, become machine level thought. Something to hold your hand as its broken down.

            @xyz i don't necessarily mean simply in the game world but possibly that would be easier to do that apps/programs to solve human needs, being they may be something newer/not done before. Basic interpreting of ideas and compiling code that works properly, from simple "human" level inputs, and materializing the idea would do wonders for those people who CANNOT think in a linear level needed for programming. Imagine what could be done by a child's imagination if we could put that into form. The out of the box thinking isn't always applied to gaming i feel bc the process has limitations of what been before. Once you learn to think a certain way, it defines your creativity from that point on. Removing the need for high level thoughts to make simple ideas come to life. Now as far as this being done with the resources available, that is the limitation to me. Possibly the power needed for such a program is just not there for the average joe to access, but that same thought is what makes the google algorithm so awesome right? It does searching on a crazy high level with minimal resources per person accessing said info. Why cant there be a similar dictionary/key relationship with human input to machine understanding. Most likely step by step guiding you towards your directive, not just do "this" and "that" happens.
            I will do my best to put an idea into words, but im not saying i have some grand answer here, just wanted to poke others brains on what they thought. Most educated individuals at the time believe flight would never "take off" (forgive my pun) and that the ONLY way that would possibly work, were via dirigibles. Not that all dreams come to fruition, the majority don't, but this level of thought/banter is essential for the advents of tomorrow.
            This is a concept for a planning/goal app that i had hoped to make at some point, very super rough iteration, but i think it gets the point across perfectly... its sort of an idea to take the habits of Tamagotchi and apply that to a goal oriented planning/tracking app.

            • Create an app that will keep track of my day to day activities via user prompts for information or any available tracking data that:
              A)creates a relevant avatar to represent each activity/data and make it look cartoony with comic style graphic effect
              B)the avatar will become bigger/smaller, darker/lighter, grow/die depending on each process they represent, relevant to level of achievement to goals.
              C)shows an "updated" background image of the avatar each time the phone opens/closed to the main screen with updated information that can load the menu whenever "action choosen by user" is done
              D)menu is a customizable user GUI to create new avatars and assign goals/info, upgrade the avatars
              "screen" environment, check specific data relevant to each activity
              E)Make a character that represents the progress of all activities, and make him become happy/mad, clean/dirty, fat/skinny depending on what each avatars data tracks. Make the character in the same graphical effect of the avatars.
              F)keep track of data in a log, that shows the actual progress, progress if all goals HAD been met, and progress if NO goals had been met and the ability to cancel or change log information for "errors" or unreported progress.
              G)create ability for user to "sell" screen space via advertisements on billboards on the main phone screen instead of having to purchase app/premium for access to all features.
              As far as getting the app into the stores and whatnot, i feel like that is a completely different issue, and on a business level of thought. The ability to run applications on different hardware is getting much easier to navigate, but that is a moving world, and i suppose to be effective in any aspect of business, its typically best to find others who specialize in services you don't have a good working knowledge of. I have little understanding of porting games into stores and making it run on different devices, but that is a great thing for the same hand holding AI to walk you thru. The steps looked pretty generic the very little i have looked into it, but require extensive knowledge on the processes. Again i may be sticking my foot in my mouth, but is it really that hard to do?
              There are totally holes in this series of instruction and that is where the ai could "walk" thru the code it thinks you want. If what it thinks you want isn't lining up, you could simply add a parameter, to have "x" or not include "y". eventually you could get come to, if nothing else, a blue print for a working app that needs the human element applied to get what you want, but that i see only as a step in progression. The functions of managing data, and processing information is nothing new and certainly there ARE programs that do this already, so the same processes can be put in place WITHOUT the programmer needing the mathematical understanding of what to do and the learned knowledge of how OTHERS have gotten to the step you want to use to produce your thought.
              Really i think the game world would be almost easier to do IF the code was allowed to be generated from the brain children of what came before you. Just plugging games into another, and using the same mechanics as a basis to build on. Is it "fair" and/or profitable are the real hindrances i see down that thought, besides the amount of computing that would be needed to understand what your asking of it. AI can read a book and describe it to you now, the google one that just release can understand up to 1 hour of a movie and explain it to you..... wtf right? that's amazing to me! why cant it get to a point to interpret code and what outcome we want and create a working concept?
              People once believe they could work harder than machines, and we have the story of JOHN HENRY, who supposedly beat a digging machine, but that time has since passed. Deep Blue versus Garry Kasparov was a great example of man "Beating machine" at chess. The next year he got beat by IBM.... and i think everyone can agree that the computer will win from here on out. This is the next digging machine/chess challenge. the old > "NO WAY A COMPUTER CAN DO WHAT I, A MIGHTY MAN CAN DO" 🙂
              OVERALL, I think that is a human aspect of every industry, to feel irreplaceable. One day i think everything we can fathom will be able to be done by machine, just maybe not in the basic/physical perspective we think currently. "How long until that happens?", is the only question in my mind personally.

              REVBENT Your responses look generated 🙂. At least separate the paragraphs for easier reading.

              Not everything that seems impossible now, should be considered possible in the future, especially if the only argument for that is merely "the power of endless progress."

              Just because some things that looked impossible in the past came into realization now, doesn't mean that everything imaginable now, will become possible in the future. It's a fallacy, and I don't think it has a name yet in the pantheon of common fallacies.

              There's the domain of genuinely not possible, as opposed to the domain of just seemingly impossible. It takes intelligence and knowledge and wisdom to discern between the two.

              To put it more poetically. We all know that Clarke said that Any sufficiently advanced technology is indistinguishable from magic. It may indeed be so. Let's accept it as a premise. However, from this it doesn't follow that any imaginable magic is possible thru sufficiently advanced technology. Yet, sadly, this is precisely what many a techno-zealot nowadays infer. It's again a fallacy. In first order logic no less.

                DaveTheCoder The y2038 threat is only 14 years away.

                I wouldn't worry too much about this. There are dozen other apocalypses scheduled between now and then.

                Jesusemora where are all the flying cars?

                Autogyros or gyrocopters have existed a long time now, most people just don't really need one.

                • xyz replied to this.