Computers started beating chess masters occasionally in the 1980s, and more consistently n the 2000s.

This year a computer has beaten one of the world's best players of a Korean game called Go; a far more complex game than chess requiring vast computing power.

My questions for the technical guys is; How long do you think it will be until a computer can beat a human at driving an F1 car around a circuit?

It's going to happen, right?

Tricky one. This machine won because it was fed 100's of 1000's of games played by humans and then analysed it.
it did not learn and evolve it's knowledge from it's own experiences of playing the game.
In my mind, this is still not true AI.
Same can be done for driving a car, but I reckon they would have to rely on GPS to get it to work, not in cockpit experience/feedback.

Select to expand quote

westozwind said..
...........................Same can be done for driving a car, but I reckon they would have to rely on GPS to get it to work, not in cockpit experience/feedback.

It'd have to be an approved GPS and worn on the helmet


BTW - I read that article/video yesterday too Chris. I think AI is limited until science can actually understand what constitutes consciousness... "i think, therefore i am"

Observing those in charge

I'd say most intelligence is artificial

Select to expand quote

westozwind said..
Tricky one. This machine won because it was fed 100's of 1000's of games played by humans and then analysed it.
it did not learn and evolve it's knowledge from it's own experiences of playing the game.
In my mind, this is still not true AI.
Same can be done for driving a car, but I reckon they would have to rely on GPS to get it to work, not in cockpit experience/feedback.

Hey I heard this story the other day. The reason this game was so special was because it wasn't fed 100's of thousands of games previously. It apparently learnt and thought for itself,. scary ****! They say its the first time a computer thought.

Haven't the people designing these things seen terminator??

Select to expand quote

AUS1111 said..
Computers started beating chess masters occasionally in the 1980s, and more consistently n the 2000s.

This year a computer has beaten one of the world's best players of a Korean game called Go; a far more complex game than chess requiring vast computing power.

My questions for the technical guys is; How long do you think it will be until a computer can beat a human at driving an F1 car around a circuit?

It's going to happen, right?

F1 is easy, but what about PWA or kiting ?Can you see waterproof robot making gybes in the chop ? or boosting kites and jumping to land head first on the sand bank ....

Chris,
We are not far away from computers that can learn very quickly there is an australian company listed on the ASX code BRN brainchip claims to have invented a chip that can function autonomously by learning as it goes .One of the first demonstrations of it was a car driving around a track 1 lap and it new its way around .the future has arrived

Select to expand quote

Macroscien said..

AUS1111 said..
Computers started beating chess masters occasionally in the 1980s, and more consistently n the 2000s.

This year a computer has beaten one of the world's best players of a Korean game called Go; a far more complex game than chess requiring vast computing power.

My questions for the technical guys is; How long do you think it will be until a computer can beat a human at driving an F1 car around a circuit?

It's going to happen, right?

F1 is easy, but what about PWA or kiting ?Can you see waterproof robot making gybes in the chop ? or boosting kites and jumping to land head first on the sand bank ....

Kiting would be easy, but windsurfing could never be mastered by AI......

Select to expand quote

Gazuki said..

westozwind said..
Tricky one. This machine won because it was fed 100's of 1000's of games played by humans and then analysed it.
it did not learn and evolve it's knowledge from it's own experiences of playing the game.
In my mind, this is still not true AI.
Same can be done for driving a car, but I reckon they would have to rely on GPS to get it to work, not in cockpit experience/feedback.

Hey I heard this story the other day. The reason this game was so special was because it wasn't fed 100's of thousands of games previously. It apparently learnt and thought for itself,. scary ****! They say its the first time a computer thought.

Haven't the people designing these things seen terminator??

Go's not that complicated. It has very simple rules and moves. it's just a strategy game of surrounding territory (board area) with stones and the one with the most territory at the end wins. I don't find it at all surprising that AI can play it from first principles.

Select to expand quote

NotWal said..

Go's not that complicated. It has very simple rules and moves. it's just a strategy game of surrounding territory (board area) with stones and the one with the most territory at the end wins. I don't find it at all surprising that AI can play it from first principles.

Have you played it? Against someone good?

The initial simplicity is deceptive. It's a very complex, chaotic game. Slight changes to the position of a piece early on have massive consequences later on.

Hi all.
Aus1111, This is bordering on a heavy weather topic... Essentially it is the ability to do the analysis based on an AI's equivalent of experience in a time frame to be useful and relevant.
The racing car scenario is already there. Each track is mapped optimised and off they go - GPS not required far more local. The more complex scenario is the random event which U deal with in every day life on the roads - new road - unpredictable surface - dodgey weather - and non-AI drivers in other vehicles.

The situation is often created where each new mountain is conquered by a specific AI. The AI is built for purpose, so one fit for a truck is unlikely to be good for underwater welding.
Ha.. my first experience of being displaced - I borrowed my brother's Pool cleaning robot - it beat me hands down the only downside is I had to give it back and am still to find the leak in my pool pipes.

It will not be long b4 a road rage video crops up of an agro human attacking an AI driven vehicle which would respond by??? I wonder.

Automated drone flight is an area getting a lot of space - there are some fascinating UTube clips - I was reading about a company publishing research on mimicking the flying ability of a bumblebee - it hits things and keeps flying - their goal was to nail that in as few clock cycles of the cpu as possible and the humble bumblebee was the model.
Although not publicised that much, the ability to be aggressive - soldier or guard type role so doing its master's bidding - the ethics of that v's the practical implications. IED's created a massive stimulus... way beyond bomb disposal.. The concept of "For the greater good" is a hot potato..

One of the mega units in AI is called Watson which since it won at Jeopardy on US tv some years back, this has shrunk to be capable of fitting in a server rack unit, and is available for hire to be customised for applications - one of which is cancer diagnosis at which it is very accurate. I am not sure what is used behind the scene's at Apple - Siri, Android - Oy Google, MSoft - Cortana but same concept..

B4 too long the issue will be judged in human consequences...basically loss of income and consequently what we rely on the gov' - loosing tax revenue. The concept of a minimum or social wage will take on a new meaning.

The take home - get good at something and stay ahead or join the company equivalent - and then there is politics!!!

Cheers
AP

Select to expand quote

WA waverider said..
Chris,
We are not far away from computers that can learn very quickly there is an australian company listed on the ASX code BRN brainchip claims to have invented a chip that can function autonomously by learning as it goes .One of the first demonstrations of it was a car driving around a track 1 lap and it new its way around .the future has arrived

The CEO is one of my best mates as it happens, which has nothing to with this thread, but yeah, very interesting stuff.

I pretend to understand when he explains it.

Select to expand quote

AUS1111 said...
I pretend to understand when he explains it.

I pretend to listen when my wife talks at me...

Depends on the definition of intelligence ?

Running algorithms and referencing data sets isn't really my definition of intelligence.

When a machine starts coming up with conspiracy theories or providing insight into the human condition I'll be impressed.

Artificial Intelligence....= Federal government ...all parties.

Select to expand quote

Macroscien said..

AUS1111 said..
Computers started beating chess masters occasionally in the 1980s, and more consistently n the 2000s.

This year a computer has beaten one of the world's best players of a Korean game called Go; a far more complex game than chess requiring vast computing power.

My questions for the technical guys is; How long do you think it will be until a computer can beat a human at driving an F1 car around a circuit?

It's going to happen, right?

F1 is easy, but what about PWA or kiting ?Can you see waterproof robot making gybes in the chop ? or boosting kites and jumping to land head first on the sand bank ....

The main prob with tea bagging is: Would AI accept to wear boardies over wet suits????

^

Select to expand quote

AI will never be able to out do humans, but it could certainly compliment humans by applying a calculating process that isn't disrupted by emotions, traits etc

Computers do that without AI.

There is a notion abroad about the coming AI "singularity" where AI reaches some kind of tipping point. It could be a bot capable of self directed learning that just takes off, making decisions, giving instructions. Some say it's just around the corner.

When your pc starts playing practical jokes you know your doom is at hand. You'll be relegated to a biological support system for your new overlord.

I'd be happy with any type of intelligence artificial or otherwise

We automatically make this connection

Intelligent = Smart

Can we envision someday Artificial Brains programmed specially to be rude, stupid , nasty, unhelpful, make mistakes all the time ???

If we want Robot that remind people , there are plenty of examples of Humans for no obvious reason being nasty , awful , vulgar and etc.

We could imagine city inhabited by smart robots, chasing each other to grab spare parts, trash each other, writing poems or painting vulgar graffitis , and rap singing songs...

I was told once in a lecture from a neuroscientist - Our brain is so smart that it knows to not allow itself to design artificial intelligence, or to allow us to become conscious of how the brain works.

Hmmmm. definitely interesting

Select to expand quote

Macroscien said..
We automatically make this connection

Intelligent = Smart

Can we envision someday Artificial Brains programmed specially to be rude, stupid , nasty, unhelpful, make mistakes all the time ???

If we want Robot that remind people , there are plenty of examples of Humans for no obvious reason being nasty , awful , vulgar and etc.

We could imagine city inhabited by smart robots, chasing each other to grab spare parts, trash each other, writing poems or painting vulgar graffitis , and rap singing songs...

Love it. LOL

It is reputed that when Albert Binet, the inventor of the IQ test, was asked "What is IQ?", he replied "It is what my test measures."

Artificial intelligence - the ability to regurgitate 'facts' learned at Uni or school......and then years later - to still believe they are 'facts'.

Aus 1111 - is far and away the most 'intelligent' person on these forums - followed closely by Rupert and Pweedas.

For once proud that I'm dumb

^ I seem to recall that we established some time ago that you're a bot. Is that an indication the AI isn't quite there yet?

Select to expand quote

Mark _australia said..
I was told once in a lecture from a neuroscientist - Our brain is so smart that it knows to not allow itself to design artificial intelligence, or to allow us to become conscious of how the brain works.

Hmmmm. definitely interesting

One day in the distant future Intelligent Robots will start asking the same question by themselves:

What is a proof that I am conscious ?

What is the proof that I am really conscious , not programmed to think that I am conscious ?

One may try this path: Humans. They consider themselves to be conscious. To become conscious I must be like a human. Behave not according to program and logic , but random, violent sometimes and unpredictable.
To test that I am am really conscious I may try to kill my friend robot. If I succeed then I should feel "guilt" as humans do.
I should feel ashamed. I may even try to hang myself .
If I succeed , it means I was really conscious. Like a humans do.
Ooooops !

Select to expand quote

Macroscien said..One day in the distant future Intelligent Robots will start asking the same question by themselves:

What is a proof that I am conscious ?

What is the proof that I am really conscious , not programmed to think that I am conscious ?

Ever seen the movie "The Thirteenth Floor"?

The real question is how long until the robots take over and make us their slaves.

Select to expand quote

Beaglebuddy said..
The real question is how long until the robots take over and make us their slaves.

we are giving the programs all our data. were basicly theyre now man

Select to expand quote

Beaglebuddy said..
The real question is how long until the robots take over and make us their slaves.

I am guessing that poor pc and laptops did so already...
How many hours each day one of us already works for them massaging their keyboard...buying them newer faster parts..feeding endless electricity...

But humans have in return: some news from the world, few porno clips, Facebook and access to bank account...

In fact very soon all this could be completely virtual and will do no difference to us, end users at all,

-If the war in Syria is real , or just created in Pixel Studio - make no difference for me as I am not going to Syria to check this out,

-Virtual Pussy Porno star could be as real as flesh and bloody , as long as on the screen only

-Facebook - all events and friends that you don't need to meet in real, or don't even have a time, as you need to spend many hours behind computer

-Bank statement online - doesn't tell you anymore how much money you have, but how much money you still owe on all credit cards, loans and mortgages ...

This week a human managed to defeat "Deepmind Alphago", the machine developed by Google that managed to beat one of worlds best players four times in a row. So the score is 4-1.

As as has been noted, the machine can do little else.

Go humans!

Select to expand quote

Macroscien said..
We automatically make this connection

Intelligent = Smart

Can we envision someday Artificial Brains programmed specially to be rude, stupid , nasty, unhelpful, make mistakes all the time ???

If we want Robot that remind people , there are plenty of examples of Humans for no obvious reason being nasty , awful , vulgar and etc.

We could imagine city inhabited by smart robots, chasing each other to grab spare parts, trash each other, writing poems or painting vulgar graffitis , and rap singing songs...

Perfect description of any Microsoft help option!