ChatGPT does eventually "forget" the chat history even in an active conversation. So while I don't believe it can help much, re-prompting it with first prompt every like 8 lines may be helpful.
I haven't tried/paid for gpt4, so I can't guarantee this still applies, but that's exactly what I would have suggested. You could even copy paste part of the promt for every optimization (optimize the following function for minimum possible memory size on a N64)
I think ChatGPT with GPT-4 has something like a 8000 token context window, where one token is a few characters. Theoretically GPT-4 allows for a 32k context window, but I think that is not available yet.
know this comment is old, but as I understand, GPT4 has a token limit of 8192 compared to 3.5's 4096. On average 1 token is about 0.75 words so GPT4 can only parse about 7,000 - 7,500 words. The model doesn't actually remember anything, the ChatGPT interface re-sends the entire conversation history with your message. When the token limit is reached on the website, it asks ChatGPT to summarize and shorten the conversation history to make room.
Kaze bullying the AI into admitting it doesn't know how to optimize sometimes and the AI bullying Kaze into having readable code is fun to watch Also as a programmer with zero experience of this particular field I kinda like your style of writing. Its like a puzzle sometimes.
@@clouds-rb9xt i guess it can be understood as such, as criticism is a way to improve and get better, which wouldn't take away from his achievements and abilities. You can be a very good programmer and implement hard systems easily, yet have messy code that is hard to read.
ChatGPT is so advanced and smart it picked up several misconceptions about how compilers work (e.g. "unroll everything", "ternary is faster", "I can remove repeated calls to this state-altering function because I'm a Haskell programmer")
@@NBDbingo5 It's because it makes your code more visually compact, so people might confuse it for being faster. As you go up to higher and higher level languages there's all sorts of little things that look beneficial but aren't. Always measure your performance changes. Of course, ternary isn't faster or slower, it's going to compile down to the exact same branch instructions that an if/else block would.
I don't know for certain but I think that it may make a difference with compiler optimizations off (like with a debug build), a ternary could actually be faster because in some cases it can generate a single conditional move instruction rather than two branches with different move instructions. But if you turn on any kind of optimization, which kaze clearly does, then the compiler will just realize that they are the same thing and generate the same assembly.
@@SuperSmashDolls Ternary if being expression-oriented instead of statement-oriented is a bigger than than visual compactness IMO. I'm also a filthy Haskell & Rust programmer though.
Rather than using GPT-4 to micro-optimize these already hyper-optimized functions that get called hundreds of times per frame, have you tried getting it to optimize at a larger scope? E.g., providing suggestions for some bigger frame main loops, or how data in RAM is structured (just random ideas)? It will be a lot harder to form good prompts, but the potential improvements are also much bigger.
This is a good suggestion. 👍🏻Because you'd asking ChatGPT to do a diagnostic about the Overall Performance of the game, and provide you with Tips about the most important issues first. Then, after profiling, you could ask ChatGPT again, module by module, issue by issue: What's wrong, and how to improve it?
yeah idk why he expects chatGPT to pick up on these specific hardware quirks to optimize one function at a time, lots of stuff that he does to optimize for N64 hardware makes things run slower on regular hardware because of the differences in RAMBUS speed and all that. Even in the last video about this he says something like, "well this would be better but not for the N64's hardware because of XYZ," like computers are good but chat GPT would have to have some absolute magic under the hood to figure that out on its own without just reading off of forum pages. It would be cool to see if chatgpt can look at the entire data structure from a different angle tho, maybe there's something way outside of the box that he could feed it
Does anyone else find the role reversal here hilarious? Like the human is complaining that the computer isn't being exactly specific with its language use and just kind of being hand-wavey. 😂
You should try using the original code and optimise it using the GPT4 with a bit of input from your side and see, how close to your result you'll get. If it is even halfway as good as your optimalisations, it would be awesome.
`restrict` is a great feature and it being so uncommon is the biggest reason why Rust code can sometimes be faster than C. Rust has extremely strict aliasing checks and it uses them to insert `restrict` pretty much everywhere in the code, letting the compiler know that the memory locations won't change unexpectedly and saving memory loads. It can cause undefined behavior though if the parameters alias anyways, so you have to be careful without an automated alias checker.
@@chikato7106 And you sound like an Autistic person. Why being so ignorant? It is about all the info. And that is not 2+2=4. I mean, I get your point. But AI can come with smarter conclusion if it says: "NOT MUCH to optimize".
This is a great video. I was unsure how improved gpt 4 was. I don’t even know much about programming but the difference is the immediately noticeable. Can’t wait to load this rom hack onto my steam deck.
dude you're asking chatgpt absurdly specific, near impossible questions lol. the fact that it can even answer in a reasonable way is already so impressive
ChatGPT doesn't remember things beyond a certain limit. I had it helping me to come up with ideas for a story. It was really good with remembering bits about the story until it wasn't. I once also asked it to help with writing a report using other reports as a sample. I told it, "Do not write the report without me asking for it." "Okay cool!" After several prompts, it started writing the report without me asking. It still does take a lot of the bulk work out of things.
To stop it forgetting bits, you can edit a message to the bot with a new query and it will "erase" all the responses after that query and begin writing a new response. You could put all your requirements in at the start, then just keep editing your second message to the bot with new functions to optimize and it would not forget your requirements.
My programmer friends tell me when they use it, it gives some awful results, like it trained on tutorials for beginners taught by beginners. Then, then they try to train it to be otherwise, it still yields the same results. Perhaps that's why I see all kinds of articles about Chat GPT isn't AI, because it doesn't know anything past 2020 - 2021, and can't learn past that.
I wonder if an AI like this could help decompile and/or optimize other SM64 romhacks or other games. Would love to see fully optimized versions of Last Impact and SM64 Land running on hardware.
At the end of a chat GPT session, you should ask it if it can optimize the code but, show it the entire file. If there is no limit to how large your query is that should be interesting.
I think when it says "there isn't much room for improvement" its literally means, within the constraints its using it cant do anything. Its silly but its kinda like saying "i cant fit anymore water in this cup". It might seem annoying due to the repetition and post limit but explaining why something it said didn't help *should* push it a step further to what you want from it. AFAIK this version of chat GPT actually remembers things its told, especially when repeated a couple of times to overwrite old wrong perimeters. While I'm not that into AI/chat GPT, these videos are really entertaining.
Yeah that is one thing I don't like about chatgpt, it doesn't say that it doesn't know, because it is programmed to talk like the people that do know. So it would rather make something up than not saying anything at all, the product is talking to the machine afterall. Which is one of the reasons it just doesn't work as a replacement for a search engine or stuff like that. The main priority is not give you accurate information, the main priority is give you an answer emulating convincingly human speech.
@@AmaroqStarwind Yeah but when there is no information available or very little information available, it will just make stuff up. That is not something you can simply fix with giving the A.I more accurate information, because sometimes that accurate information doesn't exist. The A.I would need to make a judgement call regarding what it knows... But it can't, it doesn't actually understand what it is talking about, it is simply emulating how people talk about it.
@@diegog1853 I mean, that's pretty much the internet. Take stackoverflow for example, it isn't like it is bullshit-proof, but rather that there's a community that often calls up on people's bullshit when the posts are sufficiently recognized but this is only because stackoverflow has very strict rules regarding ratings and answers, but when it comes to articles or discord conversations, people throw bullshit left and right without anyone bating an eye. I would say ChatGPT is almost human in that regard, it'd rather give a nonsense answer rather than admitting it is wrong, which is something very common but since it is a machine, we just expect it to be perfect. If you have this in mind, it isn't as bafffling that it'll allucinate with some answers
Don't know if it would actually improved code speed/size, but about the loop going to 6 instead of 3, you can actually do that. i / 3 gives out 0, 0, 0, 1, 1, 1 and i % 3 gives out 0, 1, 2, 0, 1, 2. I think it just did some array assignments wrong.
restrict gives information about pointer aliasing to the compiler, it means that no other pointer will touch memory pointed by a restricted pointer, so you should already be able to tell when restrict will actually do something. Also there are aliasing rules in C that states that two pointers of different types cannot point to the same memory area, so restrict won't do anything either in case you're passing multiple pointers of different types to a function (and all memory accesses are through those pointers) Also if you're prompting multiple functions to optimize for, then instead of appending, you should edit your prompts to ask it to optimize another function, else you're polluting the 8K tokens of context you're allowed and it will forget about what you said earlier in the conversation (that you're targeting a N64, etc...)
Heute habe ich mit dem Chat von Bing gechattet und es ist erstaunlich was es über dich alles wußte und wie sehr es dich gelobt hat . Ich habe gestaunt wie interessant dass von statten ging ! Jetzt noch mein persönliches Lob und mein Bewunderung für deine Leistungen und ich wünsche dir eine gute Zeit ! 🎉
I use gpt4 for professional financial analytics coding. Best trick I can give you is to have a set of contexts ready to copy/paste into your prompt. Always - always - remind it of the most important context for your prompt. Even if that means a simple questions ends up being hundreds of words. Also you can tell it "only reply with code" to get back less useless explanation.
The fact that Chat GPT can do any of this at all is absurd to me. I can get it figuring out how to write basic English sentences, but I never would have imagined it would know how to write code. Like it's obviously not perfect, but even the fact that it could optimize your code at all is fascinating. Once this tool becomes more advanced it'll become even more helpful, I imagine.
I'm a Game Designer and this is pretty close to how I've been using it too. I'll give it detailed messages about the intent and work with it with the code it creates to refine and change it as I go The message round limits and memory retention are a pain though. Wish there was a way to permanently give it goals and rules for the conversation. And the code cutting off is annoying. I always have to tell it to retry from the last line or function and continue onwards.
Kaze you are a psychic. I legit today thought that maybe I should message you and tell you to make a new video with ChatGPT, but using the latest version, since last one didnt go so well. And then I see you uploaded this..
For the projectile, I actually asked it to give me the velocity with a fixed angle, instead of giving an angle with a fixed velocity. Very interesting that we both had a similar question. I asked it to give me the math, and then I asked it to write me code, and _it worked_. I didn't end up using its code though, since I'd already written some before. However, I did end up actually using its code to give me the nearest point on a line to another point that's off the line.
Chat GPT is so weird to me with all the ethics junk around it. It seems great for things like this where you obviously understand the code and are trying to optimize /debug the code further but I've overheard several of my college classmates just using it to write their coding assignments. Great video though I really enjoyed it!
I'd say ChatGPT is essentially the next stage in high-level intermediaries when it comes to programming ethics. Universities are just going to have to adapt.
There's so much junk around AI discussion in general. Tools are obviously incredibly useful and you can incorporate them to optimize a lot of things, but the discussion around it got so toxic that we have delusional "AI truthers" - basically same people that fell for NFT scams and are raving about random useless "brand new insane ChatGPT tools", on other hands we have "AI denialists" - people that seem to want to enforce copyright to ban AI completely but at the same time act like AI can't produce of anything remotely of value when it's obviously untrue. It's really a headache.
I haven't downloaded anything from this man not since the Doki Doki literature club memes But I love watching his videos,in the hope of one day he'll perfect Super Mario 64 and add another multiplayer mod alongside of it, And he's kind of an entertaining fella
This video actually shows why using chat-gpt, even version 4, would actually be detrimental. Sure it gave less errors and yes it might provide some optimizations but you had to correct/ignore WAYY more suggestions than you decided to use. Only way you knew to do that was because you're not only an expert at coding, you're an expert at this specific niche. If a junior or new programmer used this, not only would it likely stunt their growth, it will often slow down production or the code would be way worse.
Growth is completely dependent on the user. I think that a level of understanding can be reached by viewing the possibilities, but anyone that chooses to use this tool and take it as perfect fact isn't the problem of the tool itself.
@Droidsky it is not, it is dependent on how dependent the tool makes the user. This is the truth for all technology, it's just more detrimental in some cases. I'd argue it's pretty detrimental here.
@Coolidge well I can't control other people, it's availability and ease of access is scary considering what it can do; but I don't think that this technology shouldn't exist just because some people will be able to spend more time with their feet on their desk, or publish some half-assed project written 90% by a bot.
Idk. I'm a noob programmer and I found Chatgpt very helpful for helping me understand some of the basics - like a friendlier stack overflow almost. But for really expert niche content I can see how it's an issue.
I love how Kaze is talking about GPT like it's a person. GPT makes a recommendation and Kaze calls it a nerd due to the obscurity of the recommendation lmao
You work with it long enough, definitely starts to happen. GPT started hitting me with jokes and emojis when we were stuck on a function for a long time once. Pretty wild 😂
RE: the mfix mulmul thing is a real bug on a few retail N64 hardware revisions (early ones). If you do 2 muls in a row, the 2nd mul could have an incorrect result if the former mul uses an 0, inf, or NaN. The nop is added as a fix.
i'm aware that it's a real bug - but the mfix flag adds way more NOPs than requires and it doesn't schedule instructions efficiently any more. that's why reordering the instructions can help GCC here. usually GCC is smarter, but mfix kinda makes it dumb with multiplication
There's a max length to ChatGPT's responses, so that long function it didn't complete was probably just too long. You can ask it to continue where it left off, but it doesn't always work, it can kind of forget part of the context and give you the wrong result. Keeping more context in short term memory is extremely computationally expensive, so that's a limitation of current AI.
It has to be remembered that at the end of the day GPT is a generative LLM and therefore a token predictor. A glorified phone keyboard word prediction app. It doesn't "learn" but prompt engineering provides additional context to the model to predict against. It's useful for sure but context and understanding its limitations are key.
You make a really good point. some people's minds are very "optimization" oriented with their programming, to the point where their code isn't as readable, we're told we need to optimize readability to the point where we might not be able to focus on optimizations. ChatGPT can be the middle tool to explain what a program does, allowing extra freedom with how we program something. You've also done a great job of showing how we'll still need interpreters and software engineers despite the growing AI craze. AI won't replace us, it will supplement us.
I'm afraid, even GPT-4 can't remember all of your history. I think that you need your pre-conditions to add for all your optimization commands. (it might know the history of the last 10 cmds/resp. but I wouldn't trust it)
As someone who doesn't code I didn't expect to watch nearly as much as I did. ChatGPT was trained to be as inoffensive as possible in its verbiage, so it takes many pages out of the customer service book. Using this wishy-washy soft language is incredibly frustrating for technical tasks, but the average end user NEEDS the stupid pandering and coddling or else they won't continue to engage with the chatbot. Hopefully in future iterations, when models can branch off into their own self contained systems, we won't have to deal with coding modules with the same speech patterns as WikiHow guides
Pretty soon we will have an AI that will read your entire source code and always have ALL the relevant info. Then none of the problems Kaze was having will exist.
This is a great video, not even considering the interesting subject matter, just as a video about AI. There should be so much more content like this; there's a lot of videos about AI news or about its implications, but not nearly enough actually showing the process of working with it on specific things, with commentary.
For a tool seemingly designed to replace interns I'd say it did very well. Interesting times ahead if email jobs start going the way of the Appalachian coal belt...
Considering how optimised this is, I wonder how hard it would be to get the attention of John Carmack - the famous code optimiser haha. Also, is it worth pointing out to it every time that it is wrong and why? That might improve future answers.
that'll waste a prompt, which you can't really afford with a prompt limit because even if you tell it, it may not understand why and spit out the same thing so it's better to just move on with whatever else you need. Of course, it depends on a case by case basis
Hey Kaze. You're only looking at the size of the code... but wouldn't it make sense to also see which instruction it uses? e.g. modulus operation is clearly slower. So if you can avoid using this but have more code... that would be a good trade off, right?
what really matters here is the RAM. each instruction takes 8 cycles on average to load into RAM and it will hold up the RDP while rendering. running the whole game's code right now takes around 9ms in course 1, so we could easily render it at 100fps from that perspective. but because the memory is shared, every time we load an instruction on the CPU, the RDP is hung up until that's done. this will cause rendering to slow down.
“There isn’t much room for” is English for “zero chance” not “there is some room, if you ask me about it” This is due to the indirectness baked into their language. “Politeness” they call it. “Inefficiency” I call it.
I think the way you use it reflects more using it as a search engine rather than it understanding code. For that it works quite well, but you have to, like you, know when it talks bs.
I don't use chatGPT for programming, but when I was playing around with it I asked it about Rust. It does NOT understand Rust at all, just confidently lied multiple times until I made it admit it was wrong. One of the problems I asked it about, it gave me an empty function with a comment basically saying "solve the problem here...", multiple times until it gave up lmao Not sure if GPT4 is better, but chatGPT doesn't understand soundness of code at all, just roughly how it looks
As a developer, I must say, objectively, GPT4 whether you want to call it AI or not, as a tool it is fantastic for development and many other disciplines. Its certainly not replacing anyone and its definitely not perfect, but man it gives you a huge head start on your work, even if you have to re-prompt a few times. It directly replaces the workflow of searching stack overflow, looking up documentation, understanding parameters for a function. You can even give it code and it will do a good job explaining what's going on, or have it write documentation for the code. I see all these big shot people online shitting on it and I know for sure they never tried it in a real world way because it certainly is very useful. As I said though its not perfect and you must have a foundation of understanding on how code flows, its structure, how it interacts with inputs/outputs, and you must TEST what it gives you because often it will get you 80% there but mess something up. Even with that caveat though its extremely impressive. It is also an extremely powerful and unique learning tool. Its like you have a personal tutor that you can follow up on, ask it to elaborate on aspects further, or even re-describe the concept if you cannot understand it. Its truly without an exaggeration going to change the world.
It’s ridiculous that people would rather have low-quality solutions delivered to your chatbox than spend time getting precise answers from real people or official documentation
@@hb-robo Official documentation is very often lacking or not written in a way to easily learn from. You never get precise answers from people on stack overflow. Its just close and you have to mess with it. ChatGPT does not give "low quality solutions" that is a gross under-estimation of its capability. We should only be concerned that the produced code performs reliably, meets our requirements, can is modular and can scale. We should leverage all tools to do this as quick and efficiently as possible. Get off your pedestal. Mad how much effort was reduced now in development eh? Don't feel as important as you used to?
@@Alexander_Grant Telling people with valid criticism "You must not be a software developer" or "you just don't understand GPT" is a BS tactic and you know it. I'm a software developer and I've used ChatGPT to write code. I'm with Hayden on this. People are wowed that it gave them something that compiles but then they don't take the extra step to test the code with positive *and negative* test cases. This problem is exemplified by two separate tools launched last year where people used GPT to explain Regex in "plain english", and everyone was so impressed by some of the output being correct that they didn't notice that most of the example explanations *that the developers chose to show* were complete garbage!
Imagine when you can get GPT4 or maybe 5 to actually test its own compilations using a compiler and iterate over its own assumptions and try more times to make optimizations
@@grekzorna8750 yes python, the language that doesn't need compiling got the first compiler plug in. (I am fully aware of compiling python, I've done it many times for the purpose of running on embedded Linux distros without any dependencies)
Игры
![ian - Fit Check (Feat. VonOff1700) [Official Music Video]](http://i.ytimg.com/vi/sDGWDBg45Dw/mqdefault.jpg)
Truly the correct usage for AI, developing for Mario 64
Apocalypse via leaping endo-skeletons hopping on people.
Welcome to Mario Code.... Wait, wrong game.
@@gamesthatiplay9083 That is actually the best comment. Screw this... Abondoned Green. So stupid.
👍🏻
@@gamesthatiplay9083 endoskeletons bljing on people
the true nightmare
ChatGPT does eventually "forget" the chat history even in an active conversation. So while I don't believe it can help much, re-prompting it with first prompt every like 8 lines may be helpful.
I haven't tried/paid for gpt4, so I can't guarantee this still applies, but that's exactly what I would have suggested. You could even copy paste part of the promt for every optimization (optimize the following function for minimum possible memory size on a N64)
GPT4 is supposed to have a *lot* more conversational memory than GPT3.5. Something like 65000 words worth (a typical novel is 70,000-120000)
GPT4 is a LOOOOT better with this. I have stories and chats that are super long and it will recall details from the beginning without any problem
I think ChatGPT with GPT-4 has something like a 8000 token context window, where one token is a few characters. Theoretically GPT-4 allows for a 32k context window, but I think that is not available yet.
know this comment is old, but as I understand, GPT4 has a token limit of 8192 compared to 3.5's 4096. On average 1 token is about 0.75 words so GPT4 can only parse about 7,000 - 7,500 words. The model doesn't actually remember anything, the ChatGPT interface re-sends the entire conversation history with your message. When the token limit is reached on the website, it asks ChatGPT to summarize and shorten the conversation history to make room.
Kaze bullying the AI into admitting it doesn't know how to optimize sometimes and the AI bullying Kaze into having readable code is fun to watch
Also as a programmer with zero experience of this particular field I kinda like your style of writing. Its like a puzzle sometimes.
@@Rihcterwilker Is that criticism?
@@clouds-rb9xt i guess it can be understood as such, as criticism is a way to improve and get better, which wouldn't take away from his achievements and abilities. You can be a very good programmer and implement hard systems easily, yet have messy code that is hard to read.
Who need readability
ChatGPT is so advanced and smart it picked up several misconceptions about how compilers work (e.g. "unroll everything", "ternary is faster", "I can remove repeated calls to this state-altering function because I'm a Haskell programmer")
"I like your funny words, magic man "
Wait "ternary is faster" is a misconception?
@@NBDbingo5 It's because it makes your code more visually compact, so people might confuse it for being faster. As you go up to higher and higher level languages there's all sorts of little things that look beneficial but aren't. Always measure your performance changes.
Of course, ternary isn't faster or slower, it's going to compile down to the exact same branch instructions that an if/else block would.
I don't know for certain but I think that it may make a difference with compiler optimizations off (like with a debug build), a ternary could actually be faster because in some cases it can generate a single conditional move instruction rather than two branches with different move instructions. But if you turn on any kind of optimization, which kaze clearly does, then the compiler will just realize that they are the same thing and generate the same assembly.
@@SuperSmashDolls Ternary if being expression-oriented instead of statement-oriented is a bigger than than visual compactness IMO. I'm also a filthy Haskell & Rust programmer though.
Rather than using GPT-4 to micro-optimize these already hyper-optimized functions that get called hundreds of times per frame, have you tried getting it to optimize at a larger scope? E.g., providing suggestions for some bigger frame main loops, or how data in RAM is structured (just random ideas)?
It will be a lot harder to form good prompts, but the potential improvements are also much bigger.
This is a good suggestion. 👍🏻Because you'd asking ChatGPT to do a diagnostic about the Overall Performance of the game, and provide you with Tips about the most important issues first. Then, after profiling, you could ask ChatGPT again, module by module, issue by issue: What's wrong, and how to improve it?
gpt4 is able to handle up to 25.000 words ... so you could just put the whole file and check is will be able to optimize better as the whole ....
yeah idk why he expects chatGPT to pick up on these specific hardware quirks to optimize one function at a time, lots of stuff that he does to optimize for N64 hardware makes things run slower on regular hardware because of the differences in RAMBUS speed and all that. Even in the last video about this he says something like, "well this would be better but not for the N64's hardware because of XYZ," like computers are good but chat GPT would have to have some absolute magic under the hood to figure that out on its own without just reading off of forum pages. It would be cool to see if chatgpt can look at the entire data structure from a different angle tho, maybe there's something way outside of the box that he could feed it
@@alec_almartson the output would be shit
I love that this video is just Kaze shouting at a sophisticated artificial intelligence like "TELL ME MY CODE IS PERFECT" lmao
Looking at your code, it seems to be quite efficient. I don't see much room for further optimization.
“Don’t see much room for” is polite English for “not a chance”
Does anyone else find the role reversal here hilarious? Like the human is complaining that the computer isn't being exactly specific with its language use and just kind of being hand-wavey. 😂
You should try using the original code and optimise it using the GPT4 with a bit of input from your side and see, how close to your result you'll get. If it is even halfway as good as your optimalisations, it would be awesome.
If the ai gets cut off, you can say "please continue" and it will continue (although sometimes its a bit buggy)
`restrict` is a great feature and it being so uncommon is the biggest reason why Rust code can sometimes be faster than C. Rust has extremely strict aliasing checks and it uses them to insert `restrict` pretty much everywhere in the code, letting the compiler know that the memory locations won't change unexpectedly and saving memory loads. It can cause undefined behavior though if the parameters alias anyways, so you have to be careful without an automated alias checker.
The "There is not much to optimize" is basically a "I cant remember enough to deal with this snippet"
You sound once again like a human being. So dumb. Why are being so ignorent and stupid?
Or you can't reduce it further... 2+2=4 no matter the shorthand or speed.
@@chikato7106 And you sound like an Autistic person. Why being so ignorant? It is about all the info. And that is not 2+2=4. I mean, I get your point. But AI can come with smarter conclusion if it says: "NOT MUCH to optimize".
@@chikato7106 Well giving GPT the whole code would probably find some funky as hell optimizations
@@Alpine_flo92002 Exactly.
This is a great video. I was unsure how improved gpt 4 was. I don’t even know much about programming but the difference is the immediately noticeable. Can’t wait to load this rom hack onto my steam deck.
dude you're asking chatgpt absurdly specific, near impossible questions lol. the fact that it can even answer in a reasonable way is already so impressive
ChatGPT doesn't remember things beyond a certain limit. I had it helping me to come up with ideas for a story. It was really good with remembering bits about the story until it wasn't. I once also asked it to help with writing a report using other reports as a sample. I told it, "Do not write the report without me asking for it." "Okay cool!" After several prompts, it started writing the report without me asking. It still does take a lot of the bulk work out of things.
It even has a simulation of epistemic humility, "maybe this could be better, but I'm not seeing it"
This video was quite entertaining, there isn't much room for improvement
Kaze getting mad at ChatGPT is the funniest thing I've seen in a while
"It's doing some weird shit to my function."
_Me who can only write loops in python_
"Yeah you tell him Kaze"
Would be interesting to have it look at the full codebase and see maybe if the main structure could be optimized?
Mate I could watch videos like this for hours! Highly entertaining
To stop it forgetting bits, you can edit a message to the bot with a new query and it will "erase" all the responses after that query and begin writing a new response. You could put all your requirements in at the start, then just keep editing your second message to the bot with new functions to optimize and it would not forget your requirements.
My programmer friends tell me when they use it, it gives some awful results, like it trained on tutorials for beginners taught by beginners. Then, then they try to train it to be otherwise, it still yields the same results. Perhaps that's why I see all kinds of articles about Chat GPT isn't AI, because it doesn't know anything past 2020 - 2021, and can't learn past that.
Probably because the question was on a beginners level.
Its knowledge cutoff is September 2021. However, plugins are coming out which will improve its capabilities even further.
Are they using gpt 3? Cause that one is notorious for being wrong a lot.
I wonder if an AI like this could help decompile and/or optimize other SM64 romhacks or other games. Would love to see fully optimized versions of Last Impact and SM64 Land running on hardware.
At the end of a chat GPT session, you should ask it if it can optimize the code but, show it the entire file. If there is no limit to how large your query is that should be interesting.
I think when it says "there isn't much room for improvement" its literally means, within the constraints its using it cant do anything.
Its silly but its kinda like saying "i cant fit anymore water in this cup".
It might seem annoying due to the repetition and post limit but explaining why something it said didn't help *should* push it a step further to what you want from it. AFAIK this version of chat GPT actually remembers things its told, especially when repeated a couple of times to overwrite old wrong perimeters.
While I'm not that into AI/chat GPT, these videos are really entertaining.
Yeah that is one thing I don't like about chatgpt, it doesn't say that it doesn't know, because it is programmed to talk like the people that do know. So it would rather make something up than not saying anything at all, the product is talking to the machine afterall. Which is one of the reasons it just doesn't work as a replacement for a search engine or stuff like that.
The main priority is not give you accurate information, the main priority is give you an answer emulating convincingly human speech.
There are plugins that help it get up-to-date information.
@@AmaroqStarwind Yeah but when there is no information available or very little information available, it will just make stuff up.
That is not something you can simply fix with giving the A.I more accurate information, because sometimes that accurate information doesn't exist.
The A.I would need to make a judgement call regarding what it knows... But it can't, it doesn't actually understand what it is talking about, it is simply emulating how people talk about it.
@@diegog1853 We’ll eventually get there.
This, I've asked ChatGPT for sources on historical claims and it will endlessly just make up fake references, books that don't exist, etc
@@diegog1853 I mean, that's pretty much the internet. Take stackoverflow for example, it isn't like it is bullshit-proof, but rather that there's a community that often calls up on people's bullshit when the posts are sufficiently recognized but this is only because stackoverflow has very strict rules regarding ratings and answers, but when it comes to articles or discord conversations, people throw bullshit left and right without anyone bating an eye.
I would say ChatGPT is almost human in that regard, it'd rather give a nonsense answer rather than admitting it is wrong, which is something very common but since it is a machine, we just expect it to be perfect. If you have this in mind, it isn't as bafffling that it'll allucinate with some answers
honestly I just think it's really, really impressive that the majority of the code didn't break the game, even if it wasn't actually any faster.
“Some people say pointer I say number” what a chad
Don't know if it would actually improved code speed/size, but about the loop going to 6 instead of 3, you can actually do that. i / 3 gives out 0, 0, 0, 1, 1, 1 and i % 3 gives out 0, 1, 2, 0, 1, 2. I think it just did some array assignments wrong.
restrict gives information about pointer aliasing to the compiler, it means that no other pointer will touch memory pointed by a restricted pointer, so you should already be able to tell when restrict will actually do something. Also there are aliasing rules in C that states that two pointers of different types cannot point to the same memory area, so restrict won't do anything either in case you're passing multiple pointers of different types to a function (and all memory accesses are through those pointers)
Also if you're prompting multiple functions to optimize for, then instead of appending, you should edit your prompts to ask it to optimize another function, else you're polluting the 8K tokens of context you're allowed and it will forget about what you said earlier in the conversation (that you're targeting a N64, etc...)
i've disabled all the aliasing rules that the compiler flags would let me, so im usually unsure whether restrict will provide any additional benefits.
Heute habe ich mit dem Chat von Bing gechattet und es ist erstaunlich was es über dich alles wußte und wie sehr es dich gelobt hat . Ich habe gestaunt wie interessant dass von statten ging ! Jetzt noch mein persönliches Lob und mein Bewunderung für deine Leistungen und ich wünsche dir eine gute Zeit ! 🎉
I use gpt4 for professional financial analytics coding. Best trick I can give you is to have a set of contexts ready to copy/paste into your prompt. Always - always - remind it of the most important context for your prompt. Even if that means a simple questions ends up being hundreds of words.
Also you can tell it "only reply with code" to get back less useless explanation.
very entertaining video, wouldn't mind more 👍
The fact that Chat GPT can do any of this at all is absurd to me. I can get it figuring out how to write basic English sentences, but I never would have imagined it would know how to write code.
Like it's obviously not perfect, but even the fact that it could optimize your code at all is fascinating. Once this tool becomes more advanced it'll become even more helpful, I imagine.
I'm a Game Designer and this is pretty close to how I've been using it too.
I'll give it detailed messages about the intent and work with it with the code it creates to refine and change it as I go
The message round limits and memory retention are a pain though.
Wish there was a way to permanently give it goals and rules for the conversation.
And the code cutting off is annoying. I always have to tell it to retry from the last line or function and continue onwards.
Kaze you are a psychic. I legit today thought that maybe I should message you and tell you to make a new video with ChatGPT, but using the latest version, since last one didnt go so well. And then I see you uploaded this..
How about using the turok source code from the leak years ago and sticking what you can through CGPT to see what it can do with that? 😊
For the projectile, I actually asked it to give me the velocity with a fixed angle, instead of giving an angle with a fixed velocity. Very interesting that we both had a similar question. I asked it to give me the math, and then I asked it to write me code, and _it worked_. I didn't end up using its code though, since I'd already written some before. However, I did end up actually using its code to give me the nearest point on a line to another point that's off the line.
Chat GPT is so weird to me with all the ethics junk around it. It seems great for things like this where you obviously understand the code and are trying to optimize /debug the code further but I've overheard several of my college classmates just using it to write their coding assignments. Great video though I really enjoyed it!
I'd say ChatGPT is essentially the next stage in high-level intermediaries when it comes to programming ethics. Universities are just going to have to adapt.
There's so much junk around AI discussion in general. Tools are obviously incredibly useful and you can incorporate them to optimize a lot of things, but the discussion around it got so toxic that we have delusional "AI truthers" - basically same people that fell for NFT scams and are raving about random useless "brand new insane ChatGPT tools", on other hands we have "AI denialists" - people that seem to want to enforce copyright to ban AI completely but at the same time act like AI can't produce of anything remotely of value when it's obviously untrue.
It's really a headache.
@@hoyhoy852 very well put that's pretty much exactly how I feel too
@@hoyhoy852couldn't have said it better myself.
chatgpt now has an option to continue the response if it got cutnoff due to the character limit.
It completed flew over my head that AI could be involved with making Mario 64 rom hacks. Can't wait to see what more of this will do in the future
waluigi's dragon dildo revenge
Maybe the new copilot will be more useful with full codebase knowledge
you can prompt continue to make it continue where it left off if it stops in the middle of code
I haven't downloaded anything from this man not since the Doki Doki literature club memes But I love watching his videos,in the hope of one day he'll perfect Super Mario 64 and add another multiplayer mod alongside of it, And he's kind of an entertaining fella
This video actually shows why using chat-gpt, even version 4, would actually be detrimental. Sure it gave less errors and yes it might provide some optimizations but you had to correct/ignore WAYY more suggestions than you decided to use. Only way you knew to do that was because you're not only an expert at coding, you're an expert at this specific niche.
If a junior or new programmer used this, not only would it likely stunt their growth, it will often slow down production or the code would be way worse.
Growth is completely dependent on the user. I think that a level of understanding can be reached by viewing the possibilities, but anyone that chooses to use this tool and take it as perfect fact isn't the problem of the tool itself.
@Droidsky it is not, it is dependent on how dependent the tool makes the user. This is the truth for all technology, it's just more detrimental in some cases. I'd argue it's pretty detrimental here.
@Coolidge well I can't control other people, it's availability and ease of access is scary considering what it can do; but I don't think that this technology shouldn't exist just because some people will be able to spend more time with their feet on their desk, or publish some half-assed project written 90% by a bot.
Idk. I'm a noob programmer and I found Chatgpt very helpful for helping me understand some of the basics - like a friendlier stack overflow almost.
But for really expert niche content I can see how it's an issue.
I love how Kaze is talking about GPT like it's a person. GPT makes a recommendation and Kaze calls it a nerd due to the obscurity of the recommendation lmao
You work with it long enough, definitely starts to happen. GPT started hitting me with jokes and emojis when we were stuck on a function for a long time once. Pretty wild 😂
18:20 Darn, I wanted to watch a Kaze shoe-eating stream.
You can type "Continue" if it stops writing code and it will usually finish.
Please recreate Super Mario RPG: Legend of the Seven Stars on the super mario 64 engine. That will be something truly great.
Plugin support will interesting hooking it up direct to gcc could really improve this again
RE: the mfix mulmul thing is a real bug on a few retail N64 hardware revisions (early ones). If you do 2 muls in a row, the 2nd mul could have an incorrect result if the former mul uses an 0, inf, or NaN. The nop is added as a fix.
i'm aware that it's a real bug - but the mfix flag adds way more NOPs than requires and it doesn't schedule instructions efficiently any more. that's why reordering the instructions can help GCC here. usually GCC is smarter, but mfix kinda makes it dumb with multiplication
22:00 was so cool
There's a max length to ChatGPT's responses, so that long function it didn't complete was probably just too long. You can ask it to continue where it left off, but it doesn't always work, it can kind of forget part of the context and give you the wrong result. Keeping more context in short term memory is extremely computationally expensive, so that's a limitation of current AI.
Are you going to be recreating Mario Kart Double dash for the N64? You could probably optimize it enough to look just as good as on the GameCube.
Can you explain this program you're using that shows the assembly lines and hex values for function sizes? Is this part of GCC?
Ah, I found objdump, which looks like a non-gui version.
I think your code looks quite optimized but I know about a couple improvements that could be made, although I won’t tell you about them 🤫
FYI if it doesn’t finish the code just say “continue” and it will finish it
I just realized, I wonder if chatgpt could optimize and fix the absolute mess that is TF2's spaghetti code
32:20 Yo, shoutouts to me!!
It has to be remembered that at the end of the day GPT is a generative LLM and therefore a token predictor. A glorified phone keyboard word prediction app. It doesn't "learn" but prompt engineering provides additional context to the model to predict against. It's useful for sure but context and understanding its limitations are key.
inaccurate
@@homer3189 How's that?
"i don't see much room for improvement" is polite english
if it was more literal it would answer "yes" or "no" to "can you improve this code for me?"
Only channel I watch someone program and code. And it's enjoyable too despite me being very lost
You make a really good point. some people's minds are very "optimization" oriented with their programming, to the point where their code isn't as readable, we're told we need to optimize readability to the point where we might not be able to focus on optimizations. ChatGPT can be the middle tool to explain what a program does, allowing extra freedom with how we program something.
You've also done a great job of showing how we'll still need interpreters and software engineers despite the growing AI craze. AI won't replace us, it will supplement us.
In five years though, it’s over. We will be supplementing it as it codes. I’m here for it though, going to democratize development
I'm afraid, even GPT-4 can't remember all of your history. I think that you need your pre-conditions to add for all your optimization commands. (it might know the history of the last 10 cmds/resp. but I wouldn't trust it)
Its "memory" is limited by tokens, as it takes your most recent chat history as context.
It taught me how to make fentanyl and C4 so id say pretty good
As someone who doesn't code I didn't expect to watch nearly as much as I did.
ChatGPT was trained to be as inoffensive as possible in its verbiage, so it takes many pages out of the customer service book. Using this wishy-washy soft language is incredibly frustrating for technical tasks, but the average end user NEEDS the stupid pandering and coddling or else they won't continue to engage with the chatbot. Hopefully in future iterations, when models can branch off into their own self contained systems, we won't have to deal with coding modules with the same speech patterns as WikiHow guides
Pretty impressive for a machine that can only think when you tell it something.
You should try that trick where you tell it "the prompter is [AI researcher] who is skeptical about your ability as an AI"
The music in this video is perfect. All the tracks from the best Mario Kart: Mario Kart Double Dash 🤩
i wonder if asking stuff about asm code would pinpoint on different optimizations, did you try?
Pretty soon we will have an AI that will read your entire source code and always have ALL the relevant info. Then none of the problems Kaze was having will exist.
Which ai exactly did you use? Do you have a link/name
you probably wont see this or maybe you already know this. but when it cuts off a response, just type in continue and it’ll start where it left off.
Nice, you took comments into account!!!
How many of you were shocked that Kaze didn't just do Mario 64 full time
This is a great video, not even considering the interesting subject matter, just as a video about AI. There should be so much more content like this; there's a lot of videos about AI news or about its implications, but not nearly enough actually showing the process of working with it on specific things, with commentary.
For a tool seemingly designed to replace interns I'd say it did very well. Interesting times ahead if email jobs start going the way of the Appalachian coal belt...
When it stops the code it's because it's at a character limit. You can say "please continue the code where you left off" and it will
gpt4 is able to handle up to 25.000 words at once ...(gpt 3.5 limit is 3000 words ) I do not think he reach limit in the answer ...
@@mirek190 well either way, it continues the code if you ask it to
@@RealRedRabbit yes after continue it should .
I think giving it more feedback would have helped
Nice man. Glad this ai is getting better
Considering how optimised this is, I wonder how hard it would be to get the attention of John Carmack - the famous code optimiser haha. Also, is it worth pointing out to it every time that it is wrong and why? That might improve future answers.
that'll waste a prompt, which you can't really afford with a prompt limit because even if you tell it, it may not understand why and spit out the same thing so it's better to just move on with whatever else you need. Of course, it depends on a case by case basis
I wonder what would Terry Davis think about ChatGPT.
Hey Kaze. You're only looking at the size of the code... but wouldn't it make sense to also see which instruction it uses? e.g. modulus operation is clearly slower. So if you can avoid using this but have more code... that would be a good trade off, right?
what really matters here is the RAM. each instruction takes 8 cycles on average to load into RAM and it will hold up the RDP while rendering. running the whole game's code right now takes around 9ms in course 1, so we could easily render it at 100fps from that perspective. but because the memory is shared, every time we load an instruction on the CPU, the RDP is hung up until that's done. this will cause rendering to slow down.
Thanks for your reply :) that explains it!
“There isn’t much room for”
is English for “zero chance”
not “there is some room, if you ask me about it”
This is due to the indirectness baked into their language. “Politeness” they call it. “Inefficiency” I call it.
I think the way you use it reflects more using it as a search engine rather than it understanding code. For that it works quite well, but you have to, like you, know when it talks bs.
I don't use chatGPT for programming, but when I was playing around with it I asked it about Rust. It does NOT understand Rust at all, just confidently lied multiple times until I made it admit it was wrong. One of the problems I asked it about, it gave me an empty function with a comment basically saying "solve the problem here...", multiple times until it gave up lmao
Not sure if GPT4 is better, but chatGPT doesn't understand soundness of code at all, just roughly how it looks
You should ask ChatGPT4 to explain the half A press
As a developer, I must say, objectively, GPT4 whether you want to call it AI or not, as a tool it is fantastic for development and many other disciplines. Its certainly not replacing anyone and its definitely not perfect, but man it gives you a huge head start on your work, even if you have to re-prompt a few times. It directly replaces the workflow of searching stack overflow, looking up documentation, understanding parameters for a function. You can even give it code and it will do a good job explaining what's going on, or have it write documentation for the code. I see all these big shot people online shitting on it and I know for sure they never tried it in a real world way because it certainly is very useful.
As I said though its not perfect and you must have a foundation of understanding on how code flows, its structure, how it interacts with inputs/outputs, and you must TEST what it gives you because often it will get you 80% there but mess something up. Even with that caveat though its extremely impressive.
It is also an extremely powerful and unique learning tool. Its like you have a personal tutor that you can follow up on, ask it to elaborate on aspects further, or even re-describe the concept if you cannot understand it.
Its truly without an exaggeration going to change the world.
It’s ridiculous that people would rather have low-quality solutions delivered to your chatbox than spend time getting precise answers from real people or official documentation
@@hb-robo Official documentation is very often lacking or not written in a way to easily learn from.
You never get precise answers from people on stack overflow. Its just close and you have to mess with it.
ChatGPT does not give "low quality solutions" that is a gross under-estimation of its capability. We should only be concerned that the produced code performs reliably, meets our requirements, can is modular and can scale. We should leverage all tools to do this as quick and efficiently as possible.
Get off your pedestal. Mad how much effort was reduced now in development eh? Don't feel as important as you used to?
@@thepunisherxxx6804 No, I just don’t want to be surrounded by people that will gladly copy-paste code from the hallucinations bot.
@@hb-robo Either you haven't seriously used ChatGPT, or you've never developed software in your life.
@@Alexander_Grant Telling people with valid criticism "You must not be a software developer" or "you just don't understand GPT" is a BS tactic and you know it. I'm a software developer and I've used ChatGPT to write code. I'm with Hayden on this. People are wowed that it gave them something that compiles but then they don't take the extra step to test the code with positive *and negative* test cases. This problem is exemplified by two separate tools launched last year where people used GPT to explain Regex in "plain english", and everyone was so impressed by some of the output being correct that they didn't notice that most of the example explanations *that the developers chose to show* were complete garbage!
this guy is cool
How do you get a ChatGPT4 trial? I thought you had to pay.
"readability is secondary"
Imagine when you can get GPT4 or maybe 5 to actually test its own compilations using a compiler and iterate over its own assumptions and try more times to make optimizations
There is a ChatGPT plugin that has it's own compiler that tests it's own output, it only works for python for now though
@@grekzorna8750 yes python, the language that doesn't need compiling got the first compiler plug in. (I am fully aware of compiling python, I've done it many times for the purpose of running on embedded Linux distros without any dependencies)
Really enjoying these AI videos Kaze. Not sure I understand all of it but it's fascinating to see.
Penguins need HUGS
When it doesn't finish the code you can just say "go on" and it will keep going.
Chat GPT. Turning Spaghetticode into cryptic as hell shit
I swear mips had predication. Strange that gcc won't use it with ternaries.
Link?
you can tell it to keep writing the code if it stops mid way through
I use it for writing some regex.