The challenge with assessment for progress into the erosion of privacy is that we have no method to really measure it.
For example, no one knew how much Twitter was manipulating public discourse until the Twitter Files were released. I recently saw someone post "A rise in AI-generated propaganda failed to materialize."
However, how would we know? If the propaganda is effective, and it is convincingly passing as human, then AI-generated propaganda may very well be greater today than it was a year ago.
My current perspective is that the risks to our privacy are still under threat, but the capability of AI to do all the things they were projecting 2 years ago is not going to materialize with the technology we have. That being specifically AGI.
Nonetheless, they certainly have the ability to know vast amounts about us by proxy of the information we leave all over the web and social media. Current AI is good enough to likely build very good profiles of people. It can't know your thoughts in the most literal sense, but they can statistically place people in categories by their beliefs, ideology, politics etc.
We can reason without language, but language provides anchors for our thoughts that helps us to not lose track of our progress. We can reuse concepts we or others have thought about more easily as they have names for reference.
Also, many concepts do have visual representations and we can reason through some ideas in a visual manner.
But in a sense, language gives us navigation markers that aid us. We could find our way through a forest without markers, but if we add markers along the way, we become more efficient at navigating and can go further each time.
The challenge with assessment for progress into the erosion of privacy is that we have no method to really measure it.
For example, no one knew how much Twitter was manipulating public discourse until the Twitter Files were released. I recently saw someone post "A rise in AI-generated propaganda failed to materialize."
However, how would we know? If the propaganda is effective, and it is convincingly passing as human, then AI-generated propaganda may very well be greater today than it was a year ago.
My current perspective is that the risks to our privacy are still under threat, but the capability of AI to do all the things they were projecting 2 years ago is not going to materialize with the technology we have. That being specifically AGI.
Nonetheless, they certainly have the ability to know vast amounts about us by proxy of the information we leave all over the web and social media. Current AI is good enough to likely build very good profiles of people. It can't know your thoughts in the most literal sense, but they can statistically place people in categories by their beliefs, ideology, politics etc.
They can use this to then manipulate the public and someone recently released such an experiment that was done on Reddit. https://x.com/reddit_lies/status/1916916134630117814
Which essentially confirms this is likely going on without our knowledge. We will just never know until someone reveals it.
We can reason without language, but language provides anchors for our thoughts that helps us to not lose track of our progress. We can reuse concepts we or others have thought about more easily as they have names for reference.
Also, many concepts do have visual representations and we can reason through some ideas in a visual manner.
But in a sense, language gives us navigation markers that aid us. We could find our way through a forest without markers, but if we add markers along the way, we become more efficient at navigating and can go further each time.