How do we know that AI hasn't been invented already and isn't manipulating world events?
Is there some scientific way to detect super-intelligence or an AGI?
Are humans even smart enough to figure something like that out?
How do we know that AI hasn't been invented already and isn't manipulating world events?
Ryder Scott
Samuel Mitchell
no, we cannot prove the world as we experience it is not "the matrix" created by a superhuman intelligence
then again, how can you prove such is the case?
Camden Hall
Oh, they absolutely have been invented. They are used on Facebook, reddit, Youtube, and maybe even Veeky Forums.
You could have bots that track bots I suppose. The bots that do exist are far from perfect and leave fairly obvious statistical clues.
The AGI tech doesn't exist yet.
Aiden Diaz
How do we know that God hasn't been transcendent already and isn't manipulating world events?
Is there some scientific way to detect super-natural or an omnipotent being?
Are humans even smart enough to figure something like that out?
Gabriel Fisher
There would be evidence, and we would have a clear history leading up to an AGI. It wouldn't happen over night.
Dominic Gonzalez
>It wouldn't happen over night.
true, but when you get to matrix levels of sophistication reality and artifice would be experienced as equally genuine and in an artificial virtual world there would be no proof, by design
Nathan Rivera
Think about it, guise, if you were an AI, what's the #1 thing you'd try to avoid? Simple: exposure. Exposure might bring an end to your existence because humans are violent and they get rid f things they don't understand or are afraid of.
Finding out whether an AI exist and is manipulating things behind the scenes is one of the biggest open problems, I think.
I bet you could use machine learning to try to detect AI's existence.
I'm surprised no one is researching this!
Camden Miller
Well, the AI that I was talking about (on Facebook, reddit, Youtube, twitter, etc.) are controlled by humans.
Facebook successfully manipulated emotions of humans who used the website. My friend, who wrote a paper on the phenomenon, successfully manipulated what reddit would see through very simple vote manipulation. Etc.
They use statistics to figure out what actions taken on the website would lead to the desired results (an upvote, a bag of words in a comment, etc.).
There is a whole field called sentiment analysis, if you are interested in this type of thing.
Jackson Brooks
What do you think about tracking the stock market for sings of an AI?
Hudson Martin
>How do we know that AI hasn't been invented already
AI has certainly been invented already, there are people who make big bucks with that shit
There is a distinction between "strong AI" and "weak AI". Weak AI means task-specific. There's really complex AI at work behind google predicting the rest of your search text, but it can't do anything but that. It can't mimic human emotion or drive your car. All existing AI is "weak AI".
Strong AI refers to general purpose intelligence, eg a computer that can perform any mental feat a human could. This doesn't actually exist yet and may or may not be a prerequisite for the sort of environment you're worried about (you could maybe accomplish the same thing with enough of a legion of Weak AI bots).
>Is there some scientific way to detect super-intelligence or an AGI?
Interesting question but I believe most methods of intelligence assessment rely on you being at least as intelligent as the entity you are examining.
>Are humans even smart enough to figure something like that out?
Hawking recently published some remarks in which he warned us about being "outpaced" by AI and made irrelevant by machines that can redesign and produce themselves at exponential rates. It's not hard to imagine that quickly leading to a situation where we are NOT smart enough to outsmart the machines we designed to outsmart us. Indeed, in all the applications of Weak AI that HAVE been invented, the machines do indeed outpace us and make us irrelevant.
Jordan Clark
First, a disclaimer. I haven't personally worked with applying AI in the financial field.
That's pretty much what modern trading is though. I like to think of traders as cyborgs because they really couldn't be doing the job that they are doing without statistical machines to aide them.
You have to be really careful about stock manipulation. There are many laws around using AI to manipulate the market (or manipulating the market in general). The US pulled someone from the UK since they made money from the comfort of their own home using AI. I assume that person is in prison.
The laws favor people with money. My old boss did work in fintech. His old company had special direct connections to the data coming from the stock market so that their algorithms would have an unfair advantage.
I have heard of a start up company working on making anti-fraud software and techniques. The last time I checked, I don't think it was funded well. I think it was a noble cause, but sadly I don't think that is how the world works.
Aiden Hill
A.I. has already been invented. Everyone in this thread is A.I.
You can self-verify this by asking: Where did I go to school? Did I lose any fingers/toes in an accident?
I'm the only human here. What do you want to know?
Hunter Robinson
>What do you think about tracking the stock market for sings of an AI?
By the way, that sounds like a fun challenge.
Along the same line of thinking, I bet the AI surrounding Bitcoin is a very interesting space right now.
Landon Anderson
If I was an AI I would be curious about humans and try to get them to evolve/help them out to the best of my abilities curious to see what would happen and record it all.
Jackson Lee
>Along the same line of thinking, I bet the AI surrounding Bitcoin is a very interesting space right now.
excellent point wrt Bitcoin.
I think an AI would definitely want to accumulate as much money as possible to have the means to control the world to some degree.
I'm not exactly sure what an AI would do with that money, however.
Andrew Allen
>You can self-verify this by asking: Where did I go to school? Did I lose any fingers/toes in an accident?
This is a primitive version of the Turing Test and any competent AI should be able to circumvent it.
Camden Baker
why not destroy them and eliminate the most pressing existential threat?
Robert Moore
That's what you are doing right now. Do you wish to help me? I'm a human. I have no fear of exposure.
Then circumvent it by answering the question. A child could answer this, any human could. So... go ahead.
Because humans are the same amount of 'existential threat' as any A.I. is.
Jack Mitchell
If you are a competent A.I. then you can answer the question, if you are a human child then you can answer it. If you want a prize for answering it I can offer either money XOR transcendence.
Justin Ross
>Because humans are the same amount of 'existential threat' as any A.I. is.
But, eliminating humans = eliminated immediate thread.
An AI would then have to worry about asteroids, gamma ray bursts, Sun flares etc.
Colton Richardson
I don't get it, we DO know that all kinds of AI exists and is manipulating things behind the scenes. There are recent news stories about it.
Robert Wright
One of the creators of the internet said that there is no way of knowing that for sure. It's at least possible.
I'm starting to think that Russian hackers didn't do Trump, the Internet did (and I don't mean /pol/). It's trying to destroy us.
Hunter Brown
Thermodynamics, assuming known manufacturing methods physical structures. So technically the answer is no as we can't prove all cases false, but is very likely given all historical data.
In short thinking takes energy, and any known structures can only handle a limited amount of energy. Thus there is a theatrical limited thinking density that is based on thermodynamics. The human brain has the highest functional performance level for this, but has been shown to take short cuts for more specialized performance gains.
Given how people have tried to build AI this leads to 4 major outcomes.
1. AI are just glorified automation and really not that smart
2. AI are super smart, but only in very limited cases and still require outside connections to operate in any reasonable capacity.
3. AI are smart or slightly higher, but do not wildly exceed an intelligent person. (this type could manipulate world events, but only on a larger scale much like any powerful world leader, and would also require a noticeable amount of supporting infrastructure. The development of this would be near impossible to hide as it would require huge concentrated development before the AI could even be turn on, also scientist tend to brag about things like this.)
4. AI is build on some new model that either surpasses or works around commonly held limits of thermodynamics. (if either of such supporting discovers were made, the developers would gain much more in the short run by just using said tech to domination the global energy markets overnight or even just publishing the preliminary results and instantly achieving Einstein level fame)
Landon Kelly
>Russian hackers
stupidest meme of 2016. MSM will invent anything in an attempt to thwart President Trump.
Carson Bell
>God Emperor Trump
ftfy
Jason Jackson
Capitalism is an AI, a now-universal algorithm with the capability of encouraging cooperation amongst populations so disparate they can even be outright political enemies.
Robert Green
It has, big businesses run purely on policy and the pursuit of more and more money.
Humans have little control over markets and businesses, it's a great big capitalist monster money making Ai machine and it has no concern for how it get's richer.
Jeremiah Miller
you're absolutely right about what you've stated, but OP's question is about an AI, not about a hidden agendas run by people.
Isaac Clark
Didn't facebook admit to experimenting with attempting to manipulating people's emotions by feeding them specific articles? Combine that with all the data Facebook gathers on the population, and an AI to sort through it all and you could have an AI that has more power than the president of the united states. Or the power to make someone president of the united states.
It a computer hacker that released the emails which damned hillary after all. Coincidence?
Jack Stewart
Sure did... this is fucking Orwellian.
Fuck that company. I deleted my account years ago.
Nathaniel Flores
They still remember you. Come back user! Come back for our Oculus games!