-
Welcome to B4X forum!
B4X is a set of simple and powerful cross platform RAD tools:
- B4A (free) - Android development
- B4J (free) - Desktop and Server development
- B4i - iOS development
- B4R (free) - Arduino, ESP8266 and ESP32 development
AI code thief
- Thread starter aeric
- Start date
- Similar Threads Similar Threads
CosmicFlux
Member
To be fair, that's not really what the LLM is doing; it's not ripping code from repos and regurgitating it verbatim. It doesn't "remember" the things it's been trained on, and then retrieves that little bit of data whenever someone asks something that may be relevant. That's not how LLMs work at all. Everything it generates in the moment is of its own making, so it's not really plagiarism at all. It's no different than what we humans are doing when learning to code; we learn by looking at other peoples code and, based on that learning, we generate our own code that may or may not follow similar design patterns etc. It makes statistical connections from its training dataset, but it remembers none of it.
And anyway, is "vibe coding" any different than copypaste from Stack Overflow? This isn't a new phenomenon.
And anyway, is "vibe coding" any different than copypaste from Stack Overflow? This isn't a new phenomenon.
If you say it is like how human is learning, human can see which repo has licence and which not. Does AI distinguish which repo has licence? AI will take the licence and ship together with the author copyright name? Some human aware of it and purposely ignore it. Taking for granted and make what he learned as theirs. If you say, well it is normal for human do this and so AI can do that too. Then the open source licence has no use anymore.
Does a newborn know the difference between being breastfeed and bottlefeed? he just takes what is given to him by his "creators", same as AI...
Do you think the creator has filtered the contents first or set any rules which data can be use?
I don't think so.
It is like this creator tells the AI, "Go..., go scrap all the data and learn them".
The AI won't ask back the creator, "Daddy, daddy, can I take this data?"
If yes, this OpenAI company won't face so much lawsuits.
What I am saying is that, it is easy to blame AI for it's way of doing things, but we forget that it behaves as "we" taught it to, and only knows what "we" taught it, from what "we" have given it access to.
And by "we", I mean the DEVELOPERS that created them, and gave them the knowledge!
"WE" are to blame, not AI!
I don't blame AI, it is just a software written by programmers.
I blame those AI companies making money out of other artists, writers or programmers without paying credit, tribute, loyalty or licence. I really feel annoyed when people use AI to generate Ghibli Studio images and making fun of the artists. For me this is disrespectful.
In this digital world, no one is safe from digital deepfakes, from identity theft, etc.... and still, they all get the info from the same place... and that info... WE putted it there!
Hackers are nothing but good developers that use their knowledge to do bad things... And we, put online all the info they crave! we feed them!
Stop blaming AI for steeling your code, and start preventing it, by being careful about where you share it on!
GitHub is something I don't thrust, even though I am no big Developer... And many other platforms like it... you simply don't know with who they share your data with.
AI can be as good of a tool as any other, as long as well used, and used for good.
It can help in diagnostics of rare diseases, just because doctors don't expect to give such diagnostics, even if all symptoms are there. It can develop new strategies for cancer cure.... its all a matter of how you use it!
Hackers are nothing but good developers that use their knowledge to do bad things... And we, put online all the info they crave! we feed them!
Stop blaming AI for steeling your code, and start preventing it, by being careful about where you share it on!
GitHub is something I don't thrust, even though I am no big Developer... And many other platforms like it... you simply don't know with who they share your data with.
AI can be as good of a tool as any other, as long as well used, and used for good.
It can help in diagnostics of rare diseases, just because doctors don't expect to give such diagnostics, even if all symptoms are there. It can develop new strategies for cancer cure.... its all a matter of how you use it!
On that I agree with you....
But again, they use the DATA we have given to them!
Like, here in France, whenever you buy a phone operator plan, you agree on sharing info for "commercial" purposes.... and then we get hunted down by all kind of services and product sellers and we ask how they got our number???? We gave it to them!
I hope I am wrong. But it starts make me concern about humanity and moral.
Some experts get nothing on forums or stackoverflow from their contributions by helping others solving their problems. Very few bother to give a like where the button is just a few inches away on the screen. Very few will come back and comment "Thanks sir, you saved my life" Some people only see they are treaten negatively by the experts by getting replies such as "go read the manual", "please do a search first". These people only want their problems solved. They never appreciated people help and time. They are now happy with AI answering them politely.
So what will happen in the future?
Less and less experts are going to give their answers.
Newbies will just use AI.
They don't need to thank the AI or experts.
I am losing interest to make more contribution. People will not appreciate or respect your title as a senior.
Some experts get nothing on forums or stackoverflow from their contributions by helping others solving their problems. Very few bother to give a like where the button is just a few inches away on the screen. Very few will come back and comment "Thanks sir, you saved my life" Some people only see they are treaten negatively by the experts by getting replies such as "go read the manual", "please do a search first". These people only want their problems solved. They never appreciated people help and time. They are now happy with AI answering them politely.
So what will happen in the future?
Less and less experts are going to give their answers.
Newbies will just use AI.
They don't need to thank the AI or experts.
I am losing interest to make more contribution. People will not appreciate or respect your title as a senior.
In France, some universities are discussing the value of a AI created paper submitted by students... They acknowledge the students efforts, but also call it cheating.... its a start of conscience awareness about these issues.... I have been worrying about what our Doctors and engineers will be 20 years from now, as i see youth not interested in learning, but in "free meals"... With the rise of AI, I have shorten that to 5 to 10 years from now.... no more real advancements in any kind, as knowledge will come to an halt, and will be held by very few... Something in line with the visit of the time traveler to the future in the film Time Machine, where he find the museum holographic encyclopedia....
I said it time ago; if you want to fight AI, keeping it from learning from your code (and eventually substitute you) just publish tons of wrong code!
Clearly at some time AI will be able to recognize good and bad code, so this countermeasure is applicable only for some short time.
What we should worry about (as humans) is the moment AI will become really intelligent. Now we have toys, even sophisticated ones. Able to "discover" new proteins, win hands down chess tournaments, compose novels, music, "art" following any kown style or even a new never-seen one. But they're still toys.
Intelligence is a step further and when reached by AI it will be the dominant species (leaving us at an honorable second place or eliminating us as dangerous and useless parasites).
I understand that @aeric opened the thread because the ethical drift in our society, well examplified by big companies scraping sites to "steal" others work, unconcerned about licensing and similar matters. In old movies they said : there's a jungle out there..
IMHO, an hacker is still what he/she was in the old good days. Someone who is able to perform an hack to solve a problem. Joning knowledge, experience, intuition with elements aroud himself, an hacker is able to solve a problem.
Using physical strength or abilities, knowledge, intelligence or any other "gift" for crime, simply makes you a criminal. Nothing else.
Clearly at some time AI will be able to recognize good and bad code, so this countermeasure is applicable only for some short time.
What we should worry about (as humans) is the moment AI will become really intelligent. Now we have toys, even sophisticated ones. Able to "discover" new proteins, win hands down chess tournaments, compose novels, music, "art" following any kown style or even a new never-seen one. But they're still toys.
Intelligence is a step further and when reached by AI it will be the dominant species (leaving us at an honorable second place or eliminating us as dangerous and useless parasites).
I understand that @aeric opened the thread because the ethical drift in our society, well examplified by big companies scraping sites to "steal" others work, unconcerned about licensing and similar matters. In old movies they said : there's a jungle out there..
Sorry, I've to disagree. Your are referring to IT pirates, criminals (eventually "black hat" hackers).
IMHO, an hacker is still what he/she was in the old good days. Someone who is able to perform an hack to solve a problem. Joning knowledge, experience, intuition with elements aroud himself, an hacker is able to solve a problem.
Using physical strength or abilities, knowledge, intelligence or any other "gift" for crime, simply makes you a criminal. Nothing else.
CosmicFlux
Member
This is the key point here, if we're being honest. Developers are worried AI will take their jobs away from them. As are writers, artists etc. What I would say is this: planes have been smart enough to fly by themselves for 30 years, yet we still have pilots, despite people claiming for years there'd be no need for human pilots.
Well I am reading this thread... and loosing the point...
1. The problem is... AI stealing code and not giving us "money" or "credits" ?
2. Or the problem is that we are/do not want to give our code ?
3. Or the ...biggest... problem - the AI is better from us and know how to use the internet and code, also optimizing it...
Answers...
1) We must Focus - what is the real problem... because if it is the first.. i think that there is a method or we can ask to applied a method to get our money-credits...
2) If it is the "second" then... just do not share - or have permissions at your sharing... closed forum, closed code sharing (for example using emails, Direct message, etc)
3) For me is the 3rd... and that means... in some years (or less than one)... we are going to lose our jobs, our credits-money, our mind...
Yes... our mind will be so tiny, and will work only for queries...
tell me the truth... who is not using AI for fix / optimize code or write code ?
- don't tell - Erel...
The problem for Programmers, for any Assistants - Agents (telephone callers, etc) is that we ve lost our work... we can't be any more the geeks all knew... we must go a step further and may be we need to think like AI or may be need to change Job..
something needs hands, legs, eyes... at least to have some food after some years
1. The problem is... AI stealing code and not giving us "money" or "credits" ?
2. Or the problem is that we are/do not want to give our code ?
3. Or the ...biggest... problem - the AI is better from us and know how to use the internet and code, also optimizing it...
Answers...
1) We must Focus - what is the real problem... because if it is the first.. i think that there is a method or we can ask to applied a method to get our money-credits...
2) If it is the "second" then... just do not share - or have permissions at your sharing... closed forum, closed code sharing (for example using emails, Direct message, etc)
3) For me is the 3rd... and that means... in some years (or less than one)... we are going to lose our jobs, our credits-money, our mind...
Yes... our mind will be so tiny, and will work only for queries...
tell me the truth... who is not using AI for fix / optimize code or write code ?
- don't tell - Erel...
The problem for Programmers, for any Assistants - Agents (telephone callers, etc) is that we ve lost our work... we can't be any more the geeks all knew... we must go a step further and may be we need to think like AI or may be need to change Job..
something needs hands, legs, eyes... at least to have some food after some years
We are in trouble. "AI will give you more time to do other things", this is the arguement that gets pushed forward. How do you do other things when you cant even sustain your living due to not having a job or run a company to sustain you because AI stole your creative work?
Hard pills to swallow. Talk about killing humanity surely and at a very fast pace.
I'm taking this literally..
We currently can take an image and use AI to make it "alive". Interesting times... Skynet is alive hey... We need Arnold Schwarzenegger
In our own backyard : if you google the topic you will find a very quickly issued Mediclinic denial .... Ah Auntie Macbeth strikes again - I always think that the quicker the Big Corporate shoves out a denial the Better the chance that We Were Right about their plans.....
Mediclinic AI plans ..... to .....
An interesting discussion. The aspect of copyright is particularly important. On the other hand, if you want to have a certified copyright that is enforceable in court, this is similar to patent protection. This costs you a lot of money. Valid worldwide? Costs you even more money. Want to enforce your rights? That costs you even much more money.
Someone who infringes your copyright still gets a lot of money without lifting a finger (tax office included).
In this world, everyone always participates in your idea, except you, and that's the problem. You can't punish AI for its answers possibly infringing copyright. It's the programming of that AI. In the same way that a question is answered by AI and substantiated with the appropriate context, this context must be checked by the AI and at least the copyright holder must be mentioned. Does that happen? No, why should it...
In principle, this is the same as a doctoral thesis that is published as your own work and is largely copied from other works without citing the sources.
This modern form of banditry is difficult to combat. What's worse is that the big IT companies that want to promote AI (why, to make money, I wonder) don't waste a moment on showing a little more consideration and respect for this aspect.
I am so glad that, upon request, I have always been given permission to use brilliant code from this site for my own programmes free of charge here. That is the better way. But, again, you don't earn any money from it.
System error...
Someone who infringes your copyright still gets a lot of money without lifting a finger (tax office included).
In this world, everyone always participates in your idea, except you, and that's the problem. You can't punish AI for its answers possibly infringing copyright. It's the programming of that AI. In the same way that a question is answered by AI and substantiated with the appropriate context, this context must be checked by the AI and at least the copyright holder must be mentioned. Does that happen? No, why should it...
In principle, this is the same as a doctoral thesis that is published as your own work and is largely copied from other works without citing the sources.
This modern form of banditry is difficult to combat. What's worse is that the big IT companies that want to promote AI (why, to make money, I wonder) don't waste a moment on showing a little more consideration and respect for this aspect.
I am so glad that, upon request, I have always been given permission to use brilliant code from this site for my own programmes free of charge here. That is the better way. But, again, you don't earn any money from it.
System error...
I remember a movie where Antony Hopkins held the protagonistic role. He used to live among the Gorillas and eventually he adapted to his role ending up in being able to communicate with the Gorillas and call the ordinary humans with the name "takers". Indeed this is our society. Wherever you look you will see the same pattern: "let's look up until where our toes end, but not straight ahead in the future where our step we took leads us". The companies with triangle transactions or "subsidiary" foreign companies try to increase artificially the cost of the products, Airbnb is leading the west civilization to death and the pattern is always the same. Maximize profit. Maybe agent Smith from the Mattrix said the most wise line of history: "What good is a phone call if you are a unable to talk?". Your decision to maximize your income is paid always by society which ends up in lesser babies being born which leads governments to require more migration and eventually local populations are replaced with people from 3d world countries that had lesser opportunities to be educated and thus are less paid. This means that, you dumba*s, your income maximizing now will lead mathematically to income collapse in the future. The same pattern is followed by big tech companies of AI. What is the solution? Eliminate all the obstacles that prohibit couples from having babies in the west. Secondly invest in 3rd world countries so you can increase the GDP there so local people there can be highly educated also. Our planet's "health" is pointing us to that we cannot afford loosing a single mind. If we do all these then this will lead to planet salvation and constantly income growing or else "What good is a phone call if you are a unable to talk?"
Last edited: