yeah, just like in the movie Ghost In The Shell: Solid State Society ... The movie was made in 2006, the setting is 2034. And something that was seen as a wild fantasy back in 2006 is eerily close to becoming reality as 2034 gets closer... The movie actually tried to be a warning about the downsides the humanity will be facing if we automate our asses off. But not everyone in the audience at the time got the memo.Now imagine facilities where people cannot resist the treatment, e.g. facilities to care for our elderly. Imagine how it will come when the workforce there gets replaced by machines - not because the machines are already perfect at the job (they never will be), but simply because it does not matter.
ChatGPT is already capable of passing a Harvard-level business school test with a B-... and, there's stuff like Udemy.com and Coursera.org for online classes... If a 5-year-old did better in a Udemy or Coursera class than a 25-year-old taking the exact same class, what does that say about the level of intelligence of the said 25-year-old?ChatGPT can only learn from the internet. Interestingly enough, we have rules and laws stopping kids from doing exactly that. Who would take advise from a 5yo that was taught by 4chan? And still people want to have said 5yo admin a server, drive a car or make political decisions. Where did natural intelligence go?
agree, It's happen today..with my cable operator,the entire support is via whastapp, at least with a human you can talk..explain yourself,but with a menu you cant hack it...is very closeIt is worse. You need to project it a bit into the future.
If you have a problem with your telco, your insurance, your supplier, and you call them, you already do not get a human to talk to and solve the problem, you get a robot to talk to.
Now imagine how that becomes when it gets clear that 99.x% of customer requests can be fulfilled by a machine (and we can easily relinquish the remaining 0.x% of customers).
We already have ISO9000, which effectively replaces quality by administrative overhead.
Now imagine facilities where people cannot resist the treatment, e.g. facilities to care for our elderly. Imagine how it will come when the workforce there gets replaced by machines - not because the machines are already perfect at the job (they never will be), but simply because it does not matter.
Nothing, I fear. We are talking business school.what does that say about the level of intelligence of the said 25-year-old?
Real intelligence failed to solve even old problems like stupidity of the general population.Nothing, I fear. We are talking business school.
Real intelligence will find a solution when faced with completely new problems, AI still fails at that AFAIK.
Yup. I studied AI in college and it's nowhere near AI. It fails at reasoning, fails at solving complex problems and fails to present accurate facts. You can ask how many genders humans have and it'll be vague. Ask how many genders dogs have and it'll say two. It's more of a Eliza Computer Therapist like in the old days except it's much more advanced. It'll be great as a next generation search engine or writing term papers based on the current data. It's not self-aware nor has the ability to learn by itself.The very fact that a human has to train this is exactly the problem.
All human biases of trainers will be transfered.
So this ain't AI in my opinion.
Its a talking bird.
Well, there have been court cases on gender determination for humans, and it's still a debate - even the very definition of 'gender' gets ppl bristling. Like, 'Men cooking while women hammer a chair together?' what about self-identification? bodily functions? emotional reactions? If an AI can consider THAT while giving a vague answer, that means that this particular implementation had a bigger and richer dataset to train on.You can ask how many genders humans have and it'll be vague.
Maybe we should stop looking for a solution to that problem and try to prove it as (un)solveable. In the mean time, go for "as good as possible, accepting to be imerfect".Real intelligence failed to solve even old problems like stupidity of the general population.
But that's exactly the usecase!You can ask how many genders humans have and it'll be vague. Ask how many genders dogs have and it'll say two.
The very fact that a human has to train this is exactly the problem.
All human biases of trainers will be transfered.
So this ain't AI in my opinion.
Its a talking bird.
Yup. I studied AI in college and it's nowhere near AI. It fails at reasoning, fails at solving complex problems and fails to present accurate facts. You can ask how many genders humans have and it'll be vague. Ask how many genders dogs have and it'll say two. It's more of a Eliza Computer Therapist like in the old days except it's much more advanced. It'll be great as a next generation search engine or writing term papers based on the current data. It's not self-aware nor has the ability to learn by itself.
I think in maybe 10 years it will be able to think like humans but again if it becomes self-aware then it's a threat to humanity. That's something Elon Musk is against.