High IQ, Artificial intelligence, and Big Blue
LoopyByteloose
Posts: 12,537
I was rather annoyed with another thread in which mention was made of an Oxford scholar that claims Artificial Intelligence would destroy the world.
But some general reading on high IQ indicates that we are quite dependent on high IQ persons to contribute to a better world, while low IQ are quite suitable for domestic work and working field crops.
http://en.wikipedia.org/wiki/Intelligence_quotient#Current_tests
I personally found Mesa to be rather boring, so I only paid dues for one year. Now I see the trend is that the higher your IQ, the cheaper the annual dues are to belong to such societies. Maybe I should have tried for a better society.
http://en.wikipedia.org/wiki/High_IQ_society
+++++++++++++
But what really annoys me is that with all this talk of artificial intelligence, I have never read of a program that could actually demonstrate a superior IQ. Do you remember Big Blue and the man versus computer chess challenges?
Why can't we just offer a world-wide challenge to AI programmers to produce a reliable software application that can demonstrate a superior IQ?
Since Parallax's motto is 'Equip Your Genius', why not have Parallax sponsor a competition? We could call it the Golden Beanie and award a prize (Cash Prize?) to the first team of five programmers to develop software that could achieve an IQ of say Three Standard Deviations beyond the norm on an array of currently used standard IQ tests. We could even increase the challenge by demanding this be down in three diverse languages (English, Chinese, and Spanish... or Japanese, Hindi or Arabic?).
In other words, until Artificial Intelligence can measure up against humans it is all a farce.
I also think that all teams that represent a company, a nation, or a university should have their sponsor match Parallax's cash prize with a similar amount tax free to all five team members.
The results would a 6X the cash that Parallax pays out. And universities such as MIT and Cambridge would be more willing to commit to the global nature of the challenge.
Of course, more detail needs to be discussed about rules. And arriving at three languages that are fair might require a bit of debate. Nonetheless, why should anyone take artificial intelligence seriously if it can't pass an IQ test?
The idea here is not to make an easy challenge that might be achieved in a year or two, but to create the Holy Grail of Artificial Intelligence programmers to be attempted many times over several years.
Parallax would merely host the prize's web page and offer the cash, while the judges and rules would be done by a panel of respected authorities in IQ testing.
Of course, if one individual could do it all, there should be an even greater reward, say 5x the cash.
But some general reading on high IQ indicates that we are quite dependent on high IQ persons to contribute to a better world, while low IQ are quite suitable for domestic work and working field crops.
http://en.wikipedia.org/wiki/Intelligence_quotient#Current_tests
I personally found Mesa to be rather boring, so I only paid dues for one year. Now I see the trend is that the higher your IQ, the cheaper the annual dues are to belong to such societies. Maybe I should have tried for a better society.
http://en.wikipedia.org/wiki/High_IQ_society
+++++++++++++
But what really annoys me is that with all this talk of artificial intelligence, I have never read of a program that could actually demonstrate a superior IQ. Do you remember Big Blue and the man versus computer chess challenges?
Why can't we just offer a world-wide challenge to AI programmers to produce a reliable software application that can demonstrate a superior IQ?
Since Parallax's motto is 'Equip Your Genius', why not have Parallax sponsor a competition? We could call it the Golden Beanie and award a prize (Cash Prize?) to the first team of five programmers to develop software that could achieve an IQ of say Three Standard Deviations beyond the norm on an array of currently used standard IQ tests. We could even increase the challenge by demanding this be down in three diverse languages (English, Chinese, and Spanish... or Japanese, Hindi or Arabic?).
In other words, until Artificial Intelligence can measure up against humans it is all a farce.
I also think that all teams that represent a company, a nation, or a university should have their sponsor match Parallax's cash prize with a similar amount tax free to all five team members.
The results would a 6X the cash that Parallax pays out. And universities such as MIT and Cambridge would be more willing to commit to the global nature of the challenge.
Of course, more detail needs to be discussed about rules. And arriving at three languages that are fair might require a bit of debate. Nonetheless, why should anyone take artificial intelligence seriously if it can't pass an IQ test?
The idea here is not to make an easy challenge that might be achieved in a year or two, but to create the Holy Grail of Artificial Intelligence programmers to be attempted many times over several years.
Parallax would merely host the prize's web page and offer the cash, while the judges and rules would be done by a panel of respected authorities in IQ testing.
Of course, if one individual could do it all, there should be an even greater reward, say 5x the cash.
Comments
The concept must feed a need of vanity. The best thing to do is to use them for the purpose
of getting us off of the planet. Resources are limited, so then, time will follow.:frown:
If it can't read a human test written in 3 languages (one of which being English) and can't pass at least the top 3 IQ tests commonly accepted these days, I don't see where there is really any merit in it.
What do we have? Google and Wikipedia. Google has subverted being informative into selling you more and more of whatever you show the faintest interest in. Wikipedia does the 'Tom Sawyer thing' of getting someone else to do the work of writing.
Not exactly the kind of AI I had hoped for.