50 Comments

  1. I think Joe Rogan just proved in this podcast that mind reading will force the population to get smarter because people have to start changing the way they think but it could also be used to manipulate us potentially, unfortunately.

  2. if you mean by "mind reading" having technology with the capacity to sense and interpret thought patterns giving our augmented selves the ability to interact naturally with our personal AI in a transparent and verifiably private manner, then yes

    if you mean similar technology with the capacity to extract data from those sensors, and from our own and our personal AIs interpretations of it, and sell that information or otherwise allow others to benefit from it without our full consent, aside from complicating the matter around questions such as "what is personally identifiable information", then no

    the gray areas of questions such as "what is personally identifiable information" are tricky and would be complicated to police within the depths of software that would perform such analysis and try to scrub the information in that case, so any legal framework around this would probably have to adhere to the principle that no means no and yes means maybe not, it depends

    how do we create useful software that actually allows people to decide what to share and what not to share?

    we don't, individuals must make those decisions for themselves, but software can perhaps make suggestions, however, do we like the way software is making suggestions now?

    can we do better, such that we're actually able to avoid boning fido?

    admittedly, it would be nice to feel like cursing out my AI, doing so because that helps me sometimes, and have it simply submit and not need to ask what it did wrong, but rather inspect my thought patterns and infer its own misinterpretations of my intentions

    I'm not really excited to allow other humans to have that capacity because I don't feel comfortable having another human being that far up my ass at the present time in my life

    one benefit of AI is that we can just switch it off without going to prison

    the true benefit of AI approaches our highest ideals and unrealistic expectations of what human beings can be but usually are not, for ourselves, from our own perspective looking out at them – it can in theory adapt to our needs more effectively than a biological organism would based on only a single lifetime of limited experiences, AI can easily exceed those limitations and we have already demonstrated as much with projects such as GPT4… I personally know exactly zero humans, myself included, who can speak as eloquently on such a variety of topics as GPT4, sad but true

    however, software suggestions are getting a little better, but that's a drop in the bucket compared to the full complexity of human behavior and intention

    we can probably come up with enough engineering tricks to get that bucket closer to being full, but it's gonna take another minute or two

    we have to understand a lot, lot more about what we're doing first

    and it's going to take ubiquitous sensing throughout our environment and sharing of that public information, legally public information, before we're even close

    AI can't very accurately predict our behavior without much, much more data, not to acceptable levels of confidence anyway

    all we have now are mere parlor tricks that suggest we need a new toothbrush or more hair gel, none of the software we have now is close to predicting our principles nor can it know when it should talk to us to get feedback and have more information before making any particular suggestion

    this last part is probably critical, our software doesn't try to consult with us before making decisions for us, it merely observes and makes guesses, likely based on biased training data for the time being, but that problem is understood a little and will improve over time like most things

Leave a Reply

© 2024 FYTube Online - FYTube.Com

Partners: Omenirea.Ro , masini in rate