Google’s DeepMind makes AI program that can learn like a human

[TulliusDetritus]

Ditto

https://www.theguardian.com/global/2017 ... ke-a-human

[JWW]

Does that mean repeating the same mistakes over and over?

[RFalvo69]

ORIGINAL: JW

Does that mean repeating the same mistakes over and over?

It would be chilling if it doesn't...

Anyway, while this is one more step towards a real AI in wargames (and other games), news like these always worry me a bit.

I remember when I was a kid and these (robots, AIs...) were dreams about the future. A future where human beings, thanks to them, would have had more time to pursue arts, philosophy, tanning on the beach... The same things were said only 20 years ago about the internet ("you work from home and you save three hours every day because you don't have to commute!")

The results, now we know, are unemployment in a growing number of sectors and stress (I send a finished job at 4PM o'clock - at 4PM and two minutes o'clock I get back an e-mail asking me if I can do "this other job by 6PM o'clock...")

I'm not saying that all these technological improvements are wonderful, useful and fascinating in many ways. Without them we couldn't have truly international communities like this very one, or a tablet with books, pictures, games, maps and a treasure trove of other functions (when I watch my daughters use their iPhones/iPads to the hilt and at high-speed - sometimes even for their studies [X(] - I always feel a bit of envy...)

However, as I wrote, what I do fear is that robots guided by even increasingly sophisticated AIs, or specialised AIs, will not give humans more free time, but just substitute them.

And, as an aside, I read a couple of books about the 2008 economic crash. While the causes were multiple and complex, when the big crisis developed, a combination of the hyper-connected world and the intervention of computers programmed to react automatically give a situation ("if X then do Y at once") increased exponentially the speed of the events. People just hadn't time to think properly, while those in charge had to take life-or-death decision after being awake for 48 hours.

Sometimes I think that the late Tom Clancy nailed it back in 1994, when in "Debt of Honor" a character said: "We must re-learn to rely on ourselves, not on our Nintendo machines..."

[E]

I'm with Stephen Hawking on this issue... a true artificial intelligence will mean the end of mankind.

[JReb]

If they come up with an AI that understands the female mind then that would be a major accomplishment!

[wings7]

ORIGINAL: JReb

If they come up with an AI that understands the female mind then that would be a major accomplishment!

Not to mention the male mind...[;)]

[ncc1701e]

ORIGINAL: JW

Does that mean repeating the same mistakes over and over?

The advantage of AI is that it does not lose its memory contrary to humans that are repeating the same mistake over and over.

In terms of strategy, I think the wall example is funny. Always in history, there are always guys that think that a wall could stop anything. See how the Maginot line did work well...

[Mobius]

ORIGINAL: E
I'm with Stephen Hawking on this issue... a true artificial intelligence will mean the end of mankind.

Thank you my AI overlord. May I have another?[:)]

[Neilster]

A funny video on this topic from the satirical website The Onion...

https://www.youtube.com/watch?v=OGxdgNJ_lZM

Cheers, Neilster

[OldSarge]

That was funny! I for one welcome our AI overlords!

[Neilster]

ORIGINAL: OldSarge

That was funny! I for one welcome our AI overlords!

"Praise to the wise ones, our robot masters" [:D]

Cheers, Neilster

[Zorch]

Remember this line from Woody Allen's 'Sleeper':

Robot Servant: The security police are surrounding the house. When will you be wanting lunch?