Do it on a blog, not 10-15 simple list, and also show how well you understand its effect, not just by simple everyday situations.
How does it for instance collect information, only via text? or can it be used in combination with headset, sounds, anything. How can AI gather information would be a huge important first part.
@RedCore@RedCore I’m not making a blog on it, sorry. I’m a pretty busy person. If I ever do I will ping you. To answer your question on data:
AI can can be trained on anything that can be represented as a number and has a pattern that it can pick up on. GPT is trained on “tokens” which are made up of the individual characters in the text in its data sets.
Image models like DALL-E are essentially trained on individual pixels in the images in its data sets.
Cause how does AI interact? is it only what I do inside a program OR can it interact with me via my computer in some way! This is curcial information and requires that people understand things at a higher level then you have just displayed, Im not trying to be little you, but I am saying there is much more to the picture than you have listed, how the AI interacts is very important to me and shoud be for everyone else
@RedCore Then it is up to you to monitor how your data is collected. In the US it is required that companies list what data is being collected. You can literally see on the Apple AppStore (for example) for each app what data it collects.
There are no restriction on AI yet and pointing to the facty China also abuses AI dosent help your case, I dont care if who does it but itt is going so fast that no one can see were we are heading, especially people who focus on the positive mostly.
We need AI experts who also are experts on computers and how everything interacts, not just the notion that it is "only" text.
@RedCore I am a computer engineer. I am in my 4th year at college STUDYING HOW COMPUTERS WORK DOWN TO THE TRANSISTOR. I also WORK IN AI. Am I not who you are looking for to explain it? I’m trying and you’re just arguing with everything I say.
I’ve literally told you at this point that you can see what data your software is collecting and how. I’ve even told you where you can do exactly that.
And that is okay with you? Have you concented to this? I havent but I never got the option. I didnt even knew they used AI and as I learned it I stopped using it
@RedCore On my computer I use Linux-based operating systems for anything other than games. Linux is FOSS, so I would be able to see if it was collecting data on me, how it was collecting data, and where.
My main phone is an iPhone, but I barely use it. If I ever needed to do something that an iPhone might maliciously collect data on, I would use my Nexus 5 instead. It’s old and android, and does not collect any data.
For social media I use #mastodon which does not collect data
So NO but you can say yes in terms of AI collecting date from our post? Seemes like a contradiction to me. I am not saying Mastodon them selves do it, but I am saying it would be easy to make a program that scans the post and inject it into an AI, very simple and I have 0 tech skills but that it my understand of it.
@RedCore Correct, and there’s probably AI models doing it! And you know what? They only have access to what I say. They don’t have access to my exact screen taps, user habits, anything like that because mastodon does not provide it to them. Instagram, facebook, Twitter, Threads, etc. all collect waaaaay more data than this app.
What about the browsers? Do you understand if they can be used to gather and collect information? After all the PIXLR.com didnt require download, only login... just saying, you're the it expert... not me, I may be abit paranoid but I dont think iI am wrong in this regared anyway
@RedCore If you look back I actually said multiple times that I don’t disagree with you. And the browser I use is Brave on most devices, and chrome whenever I use windows which, like I said, I use mainly to play games. Not sure what an AI is going to do with my Rimworld data. Come to think of it, an AI overlord training on how I run my Rimworld colony might be the worst possible outcome of AI… lol
#gpt and similar #llm ‘s can be tricked easily. There are multiple examples of this online
#ai trained on interactions on the internet has lead to racism, sexism, and other terrible isms being developed by the model in the past.
When we reach #gai / #agi it could do anything. We have no way of knowing what it could do, no matter what anyone says. One thing is for sure: it will know us better than we know ourselves very quickly.
Add comment