The thing that bothers me the most about this is that we all provide the data for the AI model. We do this without knowing and without compensation, then we are sold the thing that we essentially provided the fundamental data for. I highly recommend searching for "Data Dignity" and Jaron Lainer.
Literally every service ever you sign up for requires to you accept all sorts of legally binding documents usually referred as an EULA, Terms of Service & Privacy policy so I don't see how you provide data without knowing it? Don't you read those documents? You are the product and you give the consent and these big companies anonymize the datasets before they're used in training. If you cannot identify who's data is related to those floating point numbers that these models literally are, then who cares? It's not like your name, date of birth and address are in human readable format inside the machine learning model. Personally I wouldn't care if snippets of my code end up in some other dudes project. I "steal" code from stackoverflow all the time and I have yet to hear any legal issues related to that practice, everyone does it. I've been running AI assistants for 2 years now and they have definitely made my life a lot easier. People will figure out how to use these the right way and won't lose their jobs any time soon. They just have to adapt.
Jaron Lanier is right. AI is theft. We should organize on a legal global level to fight against it in the sense that our data should not be free. Everyone in the world should be compensated through microtransaction on every bit of data
My issue: 1) The AI takes input and it got to be trained with mass of data. This input is based on OpenSource or OpenKnowledge ...and it takes this with no kind of credit or preserving copyrights or paying this source in some kind of way. And then it turns this OpenSource/OpenKnowledge by processing and generation its output into some kind of proprietary, centralized, commercial product. ChatGPT doesn't tell you on which base and which references it used to give you the answer. 2) People who get used to ChatGPT may loose the grip to the ground on what they are doing. Maybe you will once not be able to judge, if the AI is wrong, but use it anyway, because it's too complex to understand in one lifetime. What will happen, if the AI is generating some code, which is used but not understood, and got some vulnerabilities or backdoors...
Thank you for the great video! I just checked the ChatGpT and as a Junior software developer I was a bit scared when I saw that it can write Java functions:) but you are right, I don’t think AI can replace the whole context for now
I still think AI is a bubble\hype. The engineers(Tesla, Google, GM, Wayve) working on level 5 autonomy for driverless vehicles already know this. But its nice to see stuff like ChatGPT which could possible help blog writers and young artists.
You're wrong about that. It's a matter of usable data. Just as the development of car safety features was an incremental process so is the self driving car problem. They wouldn't market the technology if it was completely impossible because it's not. It's just a matter of time until every possible scenario that could happen has happened and the companies have the data for it. Why you think Tesla, for example, already has built all the necessary things to control every aspect of the car with computer. Because only thing they lack is safe model for self driving and once they have it, it's only software update and $$$ away.
yep, the only that are close are the one from Google and that's a very controlled environment. i will wait for this, just because something advanced really fast it doesn't mean it will keep its pace. In general the technology that surprises us the most are the ones that we didn't expect
The thing that AI can't necessarily replace is human interpretation (until it becomes sentient lol), so it's still gonna fall short when it comes to art imo. I mean it's probably good enough for UI, bland aesthetics, textures and corporate speak, but it won't connect with you as someone interpreting their lived experience into a song, painting, etc. You can iterate through stuff I'm sure and come up with something original perhaps, I've got a buddy who is paralyzed that uses it so it's definitely nice to see the accessibility. But I also think that things like any games that fire all their writers for AI are going to be pretty easy to spot. Just because it can take jobs doesn't mean it will do well, especially when it will be limited by the aims of corporations looking to replace jobs with a sanitized version of the human experience.
It will certainly take out jobs from humans. But in most of the cases that will be low or medium skilled ones. I'd say if you're expert in your realm, you shouldn't care much, but if you're just a beginner or decided to start a new career somewhere in arts, writing or programming - think twice.
I understood some of your words... On the topic, I doubt this AI will manage to do anything useful in my field, which is physics modeling and simulation.
Except latest breakthroughs in various fusion energy technologies wouldn't have been possible without ML models that predicts plasma flows inside the reactors among other things. So yes they are coming for you too. There is not a industry that doesn't get affected by these tech.
With the emergence of an AI writer, the next logical step will be to create an AI reader, because no one in their right mind is going to read that nonsense. And then we'll let them talk to each other
everybody that's talking about this is putting a positive spin on it. i think it's blatantly obvious that within 20 years time humans will be completely obsolete in many jobs, especially things like programming. programmers are already using chatgpt to aid in their workflow, so it probably has taken away jobs already.
If it can replace lawyers, the world will be a much better place
not sure u thought that one through
@@Ryan-xq3klyeah ai endlessly filing lawsuits sounds like a nightmare
The thing that bothers me the most about this is that we all provide the data for the AI model. We do this without knowing and without compensation, then we are sold the thing that we essentially provided the fundamental data for. I highly recommend searching for "Data Dignity" and Jaron Lainer.
Very true. It’s Google 2.0.
Literally every service ever you sign up for requires to you accept all sorts of legally binding documents usually referred as an EULA, Terms of Service & Privacy policy so I don't see how you provide data without knowing it? Don't you read those documents? You are the product and you give the consent and these big companies anonymize the datasets before they're used in training.
If you cannot identify who's data is related to those floating point numbers that these models literally are, then who cares? It's not like your name, date of birth and address are in human readable format inside the machine learning model.
Personally I wouldn't care if snippets of my code end up in some other dudes project. I "steal" code from stackoverflow all the time and I have yet to hear any legal issues related to that practice, everyone does it. I've been running AI assistants for 2 years now and they have definitely made my life a lot easier. People will figure out how to use these the right way and won't lose their jobs any time soon. They just have to adapt.
Jaron Lanier is right. AI is theft. We should organize on a legal global level to fight against it in the sense that our data should not be free. Everyone in the world should be compensated through microtransaction on every bit of data
My issue:
1) The AI takes input and it got to be trained with mass of data. This input is based on OpenSource or OpenKnowledge ...and it takes this with no kind of credit or preserving copyrights or paying this source in some kind of way.
And then it turns this OpenSource/OpenKnowledge by processing and generation its output into some kind of proprietary, centralized, commercial product. ChatGPT doesn't tell you on which base and which references it used to give you the answer.
2) People who get used to ChatGPT may loose the grip to the ground on what they are doing. Maybe you will once not be able to judge, if the AI is wrong, but use it anyway, because it's too complex to understand in one lifetime.
What will happen, if the AI is generating some code, which is used but not understood, and got some vulnerabilities or backdoors...
Thank you for the great video! I just checked the ChatGpT and as a Junior software developer I was a bit scared when I saw that it can write Java functions:) but you are right, I don’t think AI can replace the whole context for now
That is an excellent point of view bro, I don't know if my English is improving, but I got what you think about IA and I totally agree with you
Well said...
I still think AI is a bubble\hype. The engineers(Tesla, Google, GM, Wayve) working on level 5 autonomy for driverless vehicles already know this. But its nice to see stuff like ChatGPT which could possible help blog writers and young artists.
You're wrong about that. It's a matter of usable data. Just as the development of car safety features was an incremental process so is the self driving car problem. They wouldn't market the technology if it was completely impossible because it's not. It's just a matter of time until every possible scenario that could happen has happened and the companies have the data for it. Why you think Tesla, for example, already has built all the necessary things to control every aspect of the car with computer. Because only thing they lack is safe model for self driving and once they have it, it's only software update and $$$ away.
yep, the only that are close are the one from Google and that's a very controlled environment. i will wait for this, just because something advanced really fast it doesn't mean it will keep its pace. In general the technology that surprises us the most are the ones that we didn't expect
@@victoriap1561 You nailed it!
Hey Ryan I was wondering if there were any blogs or newletters you subscribe to or read for tech news?
Software especially system design is creative
Great video as always. By the way, have you done any interesting Blockchain projects recently? Really enjoyed your old ones.
I have a few topics I can talk about, will consider it
it'd be cool if the Chat could just take over the damn meetings so I could code away :D
Yeah
dangmattsmith as a developer fr
Cool! I learn English from you bro.
The thing that AI can't necessarily replace is human interpretation (until it becomes sentient lol), so it's still gonna fall short when it comes to art imo.
I mean it's probably good enough for UI, bland aesthetics, textures and corporate speak, but it won't connect with you as someone interpreting their lived experience into a song, painting, etc. You can iterate through stuff I'm sure and come up with something original perhaps, I've got a buddy who is paralyzed that uses it so it's definitely nice to see the accessibility. But I also think that things like any games that fire all their writers for AI are going to be pretty easy to spot.
Just because it can take jobs doesn't mean it will do well, especially when it will be limited by the aims of corporations looking to replace jobs with a sanitized version of the human experience.
No, I don't think so, not yet TBH, its amazing as a resource to help with things.
It will certainly take out jobs from humans. But in most of the cases that will be low or medium skilled ones. I'd say if you're expert in your realm, you shouldn't care much, but if you're just a beginner or decided to start a new career somewhere in arts, writing or programming - think twice.
I personally think sometimes writing is more artistic than arts. Blog writers r not writers I’m talking about 😊
I understood some of your words...
On the topic, I doubt this AI will manage to do anything useful in my field, which is physics modeling and simulation.
Except latest breakthroughs in various fusion energy technologies wouldn't have been possible without ML models that predicts plasma flows inside the reactors among other things. So yes they are coming for you too. There is not a industry that doesn't get affected by these tech.
@@plumbingphasejust wait until a physical robot that's equipped with real life logic for performing actions and physical tasks😅
It will but not by ChatGPT!
I want to get into software dev, but if AI steals that from me.. there’s also dangerous deep sea fishing for danger pay.. 😂
Bro it'll be like Mass Effect. Tech all around but AI is going to be banned across the board following the first global machine uprising.
With the emergence of an AI writer, the next logical step will be to create an AI reader, because no one in their right mind is going to read that nonsense. And then we'll let them talk to each other
everybody that's talking about this is putting a positive spin on it. i think it's blatantly obvious that within 20 years time humans will be completely obsolete in many jobs, especially things like programming. programmers are already using chatgpt to aid in their workflow, so it probably has taken away jobs already.