We may not be masters of our planet for very long...
Not if the T800 is anything to say about it! But strangely enough AI domination
of our entire planet has actually become a really topical issue over the last
number of years. People on many sides are saying that computers and artificial
intelligence are the biggest threat to humanity. Yeah, yeah but your scientists
were so preoccupied with whether or not they could they do start to think of
this should. And this culminated in the slightly scary announcement that
Facebook had to shut down its artificial intelligence robots after they started
communicating with each other in a secret language, which as news sources
would have you believe is the first step in global domination. But more on that in
a minute, first let me go to a previous debate between two of the biggest tech
innovators in the world Mark Zuckerberg and Elon Musk. They recently had a debate
over artificial intelligence. Mark Zuckerberg of Facebook sees it as
something that will benefit all of humanity. Elon Musk of SpaceX and Tesla,
yeah not so much. It began when Zuckerberg posted on Facebook sharing
his optimism of the future of AI and how could lead to huge breakthroughs now
most recently said that AI is the greatest risk we face as a civilization
and so Mark Zuckerberg believes in his own words that people who are fearful of
AI are pretty irresponsible most replied that Zuckerberg knowledge of this area
was limited to burn as the heads of two massive companies both of which
inherently rely on computer intelligence this brings up a debate we really should
have been having for the last ten years and it's only finally entering
mainstream culture now a couple of years ago Elon Musk pledged ten million
dollars in grants for academics to investigate AI safety so how safe are we
and how far away is true artificial intelligence the two Facebook AI systems
they have to be shut down because they were talking to each other in language
no one could understand or rather the
researchers conducting the experiment they couldn't understand it
researchers at Facebook artificial intelligence research built a chaff lot
earlier this year that was meant to learn how to negotiate by mimicking
human trading and bartering but when the social network paired two of the
programs against each other nicknamed Alice and Bob by the way they
started to develop their own bizarre form of communication to negotiate the
bartering the chat bot conversation led to divergence from human language as the
agents developed their own language for negotiating know the bots never did
anything more nefarious than discussing with each other how to split an array of
given items into a mutually agreeable split the intent was to develop a chat
bot which could learn from human interaction so fluently that an end user
could talk with arroba and not even realize it the company chose to shut
banned the chat BOTS because they said our interest was having BOTS who could
talk to people but what's really freaky is that the chat BOTS started to learn
how to negotiate in really human ways they would for example pretend to be
interested in one specific item and then later pretend they were making a big
sacrifice by giving it up AI is getting scarily real whether you believe that to
be a good thing or a bad thing Stephen Hawking Elon Musk and Bill Gates
have all commenced saying that AI could pose a threat to humanity unlike any
we've seen before if we don't keep a handle on it true artificial
intelligence if it's if it's realized could pose a threat to humanity unlike
any previous technology including nuclear weapons and if its development
is not managed carefully humanity risks engineering its own extinction central
to this is the prospect of what's called an intelligence explosion written about
by philosopher Nick Bostrom in this scenario AI gains the ability to improve
itself and exceeds the intellectual potential of the human brain basically
AI will keep improving itself at such a rate that humanity will just never be
able to catch up with it suppose for example
there was an AI with the IQ of over 6,000 what would we do then however the
people who say that artificial intelligence will not be a problem they
tend to be the people who work with artificial intelligence so I don't know
if they just know more than we do they're trying to protect it or what's
most likely they're already being controlled by a computer many
researchers regard the pessimistic view of the future as implausible or as a
distraction from the near-term benefits and moral dilemmas posed by the
technology not least because AI systems today can barely guide robots to open
doors but to be fair they can do many other things like speech recognition
facial recognition and computers now can even mimic actual people so well that
humans won't be able to know the difference for example look at this
software what if there was an AI controlling this and that was broadcast
on all TV screens and computer screens that could start wars
so how far away is actual artificial intelligence well in 1951 Alan Turing
argued that at some point computers would probably exceed the intellectual
capacity of their inventors and that therefore we should have to expect the
machines to take control whether this would be a good thing or a bad thing
he didn't say six years later Herbert Simon another researcher
declared that machines would achieve human intelligence in the visible future
the crossing of such a threshold he suspected could be psychologically
crushing but he was on the whole optimistic we must also remain sensitive
to the need to keep the computers goals attuned with our own he later said but
added I'm not convinced that this will be difficult for other computer pioneers
the future appears more ambivalent Norbert Wiener the father of cybernetics
argued that it would be difficult to manage powerful computers or even to
accurately predict their behavior complete subservience and complete
intelligence do not go together the future will be an even more demanding
struggle against the limitations of our own intelligence rather than us hanging
out on hammocks being waited upon by our robot slaves and regarding AI
development pretty much everybody sees progress they never expected here we
have a YouTube clip of a four-legged robot one of its designers tries to kick
it over but up quickly or gains its balance scrambling with uncanny
naturalness one researcher said a problem that had
been viewed as very difficult where progress was slow and incremental was
all of a sudden done locomotion in an array of fields speech processing facial
recognition language translation the approach was the same researchers
working on computer vision had spent years to get systems to identify objects
in almost no time the deep learning networks crushed their records in one
common test using a database call imagenet humans identify photographs
with a five percent error rate Google's Network operates at four point eight a I
systems can differentiate a Pembroke Welsh Corgi from a Cardigan Welsh Corgi
which even I can't do the computers are literally better at recognizing images
than us so the future of AI it's a really exciting thought or a terrifying
one let me know what you think below thanks for watching Mike ed
No comments:
Post a Comment