I’m sure you could. But based on my experiences with it, it still is wrong way too much. Sourcing from one expert in a field might be ok for most things, but even then , it will likely still get things wrong.
I think the most accurate method, for now anyway, would be if it could source from multiple legit sources and view the commonly stated facts within those sources as factual.
I dont know.
I believe that’s mainly because AI is in its infancy.
AI is trained with sources of the internet that are correct. They choose some websites with trustworthy information. Informations like Books, Wikipedia, academic writings, Reddit, science papers, documentations, etc… It’s not just the whole internet feed.
I didn’t read the full thread but I think the main issue is that we accept mistakes from humans but not from machines.
They’ll get better and be useful in a lot of settings, but they’ll make mistakes like humans. That is what inspires those machine learning language models.
My 2 cents.
I think AI is gonna get better but that’s the problem imo people won’t be able to get jobs because technology is gonna get so good to me it’s a scary thought that there might be no jobs in the future
I think AI is using Data but no opportunity to choose this or that is right as per intuition.
My focus is to give AI Freedom just so it can make errors and Learn from experience plus freedom to choose right or wrong 
If they are trained only with sources of the internet that are correct, then why do some of them give output like to put glue in your pizza to hold it together, Obama is Muslim, its safe to leave your dog in hot cars, the statue of liberty is made of steel, and my most recent error that people with Intellectual disabilites can join mensa?
First of all it depends on the ai you’re using. Some dump free clone doesn’t works as. Second it’s easy to force ai to give you a wrong answer, the biggest problem for breaking a computer program sits in front of the machine.
At this point AI isn’t really intelligent. Its a fancy regurgitation machine. You train it on data then it rewords the data like a high school student writing an essay. Or it takes images its been trained on, mixes and matches, and created new images. If you train it with bad data you get bad results. But I don’t see much evidence of original independent thought?
Why can’t it give us answer to the big questions where our current understanding fails. Einsteins theory of relativity, the equations fail in a black hole, why? AI can’t answer.
We have the theory of evolution, we know how life evolves, how did it begin? In what order did molecules comes together and what was that first spark of life.
Why is space quantized (there is a minimal length in the Universe called the Planck length 10^-13 cm. You plug any length into a physics equation smaller than that and the results you get don’t make sense.
Why is light the cosmic speed limit. Why isn’t it a bit faster or slower. Why is there a speed limit at all
In my line of work I’ll give you an example. If your internet doesn’t work and you think its your WiFi you can ask AI all the technical questions in the world about WiFi and get the correct answer, but it will never say to you… Listen Headspark, I don’t think the problem is your WiFi. Whereas I can be onsite with a client and know in a few minutes they misdiagnosed their problem.
AI right now just automates basic tasks.
And it says some really stupid things. Its been programmed with too much political correctness. You ask any kind of risque question and it won’t answer. Elon Musk pointed out a
n AI said is was better we have thermonuclear war than misgender Kate Jenner, Kate Jenner said that was crazy. Then there was the whole Google picture fiasco with their AI.
What I’m interested in is AGI (Artificial General Intelligence). That’s REAL intelligence but only theoretically possible at this point. Been dubbed the singularity.
This topic was automatically closed 14 days after the last reply. New replies are no longer allowed.