Frogboy

Put bluntly, developing AI is a big mistake.  The consequences, which will be broad and far reaching, will be horrendous, impacting negatively on humanity across the board, not that the people.... er, fools developing it care.  Just because they can develop AI doesn't mean they should.

How do we know this is actually you posting these articles and not some script Paul Tozour wrote?

That may be a scenario that could play itself out, but there are other, more immediate concerns.  Mankind is lethargic and lazy enough with the tech available to it now, so having AIs doing so much more, practically wiping its arse, mankind will become decidedly lazier.  Perhaps then, the machines will see mankind as obsolete and using valuable resources without contributing anywhere near enough to warrant its existence.

So yeah, a 'Terminator' situation may well come to pass.

My biggest concern regarding AI is  - who will be held responsible if something goes wrong.

Can't make an AI (no matter how sophisticated ) legally  liable for it's mistakes.

Corporations are people too, so the executives of said corp aren't legally liable.

The shareholders/investors are passive investors, so they can't be held liable.

It probably took dozens if not hundreds of programmers to build the AI and millions of lines of code. So there is no way to pinpoint the line(s) of code that caused the problem, and no way to know who wrote those line(s).

This leaves the corporation, but the only way to punish it is to assess a fine..but not too big a fine, as that would cost jobs.... And a small fine is just written off as the cost of doing business, which encourages repeat behavior.

Can't regulate AI, as the anti-regulation zealots won't let any regs pass.

And it's only a matter of time before SCOTUS makes AI people too.

So in the end, it's the consumer who gets screwed.

By the time an AI can replace my job, I'll either be retired or dead.

Right now, AI's are idiot savants. They are really good at only one thing. You can't take over the world by being an expert at, chess, Jeopardy, medical diagnosis, weather prediction, or road navigation.

How is this any different than, say, car brakes?

Also, if they make AI people, they'll be eligible for welfare, which means someone will produce them by the millions just to collect the checks   it'll be fucking hilarious!

You raise one of the basic issues of our civilization: the impact of AI.  You then go to a secondary issue, the sustainability of a 7 billion human population where the continual cycle of replacing human labor with machine labor intensifies.  This cycle really got got going with the industrial revolution.  There many workers lost their jobs to 'more efficient' machines.  They threw monkey wrenches into the gears of the machines in a vain attempt to prevent losing their jobs.  The humans lost.  Since this beginning, we have witnessed successive waves of automation replacing human workers.  Two aspects of this deserve consideration. 

1) First, back in the day when movies (cinema) were a new form, the German scene produced movies like Metropolis.  They envisioned a future where machines had freed humans from daily drudgery.  These humans lived to create as they wished (self-actualization?).  Underlying this was the idea that all the natural resources used to accumulate the wealth needed to produce the machines actually belonged to all the people, not just the business magnate. [cf. All citizens of Alaska get a piece of the oil revnue as their right.] The American version of this vision was "Thine alabaster cities gleam, undimmed by human tears." Charlie Chaplain explored the not so nice aspects of these Modern Times.  This displacement of workers happens over and over again - but only to a small segment of society at any one time.  So its easy for the 'better off' folks to not really care enough to do anything significant to ameliorate the process.  In fact, the dispossessed folks are often stigmatized. Lazy leeches!

2) Second, how serious is the threat of AI to humans?  This theme has been explored numerous times, with great insight.  I, Robot (the newer movie) shows how even an AI bound by the the 'three laws' ends up enslaving the human race.  [The three laws were written by Issac Asimov - seemingly a guaranteed protection against an AI run amok.] I, Robot showed just how badly it might go.  The other exploration of the theme of AI vs. biological sentient beings was masterfully explored in the Battlestar Galactica re-imagined series.  (NOT the first series which was Ponderosa in space, lol.)

3) Conclusion:  What I see as the likely result of creating an AI that is programmed to improve itself will be disastrous for our great grand children.  AI can evolve millions of times faster than multi cellular biological creatures (we humans).  They certainly can replicate faster.  So, why would we create the instrument of our eventual demise?  Simple, short term, someone will gain an advantage (mustard gas)(A-bomb)(napalm)(vulture capitalism)(cutter bombs)(etc.).  And because we can.

The Swiss always vote these kinds of social referendums down the first 2 or 3 times, before they approve them.  It takes time to raise sufficient awareness and educate people.

With automation coming to make most of humanity simply unemployable (which is why we designed and built them, after all), a Basic Income Guarantee is the only workable economic solution going forward.  

Unless you're an anarchist, of course.  

Well, unless we're going to let 75% of the population starve and die simply because we have no work for them to do anymore, it will have to, one day.

Eventually we will all be Paul Tozour.  So it matters not.

Indeed.  That was the estimate (not part of the actual referendum) because of the high cost of living in Switzerland.  But the number would vary from country to country.  The Swiss referendum was deliberately left vague because it was about raising awareness this time around.  No one expected it to pass the first try.

It's sad that society is already arguing about how much displaced workers may receive.  It's sad because the powers that be, governments and corporations, have plans in motion to displace even more with increased automation and computerisation, plus the increased taxes with fewer payees contributing to national revenue.

Even worse still, when those government backed corporations manage to create sentient AI.  Combined with sophisticated automation... robotics, sentient AI will very nearly replace the human workforce, leaving millions of displaced workers and their families living well below the poverty line on meagre handouts and little or nothing constructive to do.  Chaos and anarchy will ensue... impoverished people with idle hands will seek ways to occupy their time as they struggle to feed and clothe themselves. 

As the say, the Devil makes use of idle hands, and the crime rate will rise significantly as the displaced, and most likely homeless, fight to feed themselves and find shelter.  It is a recipe for disaster, not that the corporations will care one iota.  They'll be up there in their high security ivory towers looking down without a worry.  The automated, robotic and sentient workforce will continue producing and increasing profits, so the rabble below is irrelevant, superfluous to their needs.

Unfortunately, however, the new masters will witness the same chaos and anarchy... and by then the creators will also be seen as irrelevant and unnecessary, ushering in a scenario alike or similar to Skynet, with the extermination of humanity of the Terminator movies. 

It can't happen?  Think again!  The greater the sentience, the greater the capacity to reason such things AND take action to eliminate the threat... and given its mentor/teacher/creator is of mankind, a species capable of mass murder, the atrocities of war, genocide, the 'new master' would have little or no reason to spare what has become a somewhat pitiful species.  I mean, really, why would it?

Just because a sentient AI is supposedly possible, it doesn't mean we should create it.... and anyone with half a brain would see that we shouldn't.

Assuming AI is the best thing ever is a huge mistake.

But writing one is not that terrible complicated if the computing power  and devellopment theme one has is big.Mean using cameras to detect movements and lines/edges  is already possible, motors that can move in any possible direction too.Fully Simulating an half blind eye in real time  for an  AI will be possible in the next years. Analyzing Syntax of language google is already working on.Voice recognition and optical character recognition has improved terribly in the past years.

It will take some years but not an endless time.

I think we are a long way from the point where we can start debating over artificial -intelligence- & artificial -sentience- where a robot apocalypse type situation could be an issue.  With that said, we are very much on the verge of some very unpleasant times as we approach a situation where billions of people literally could not possibly add value to the economy.

 While the fast food ordering kiosks & self checkout register type things have been quietly progressing for years, they are just part of a larger trend.  things like the gourmet to order burger making robot, amazon warehouse automation, amazon drone delivery, IBM's purchase of weathercompany but not weather channel, adidas' announcing they were making localized fully automated factories, and the like are the tip of that unpleasant iceberg.... but I don't think it will really hit home & sink in for most people until self driving cars/trucks are commercially available

Take warehouse jobs, years ago they were often considered to be pretty decent paying jobs, the city my mom lives in not too long ago gave amazon a sweet multi-year/multi-decade tax free deal for putting their warehouse there... they did not realize however that it would be staffed almost  entirely by robots & low wage temp contract workers till more recently though. Ask most people what they think about self driving cars, they will probably talk about how much work/relaxing/etc they could get done letting the car drive instead of dealing with idiot drivers... but that misses the larger picture.  Specifically there are millions of people in the us (and more around the world) that make pretty good middle class incomes (or come close) by driving everything from taxi cabs, airport shuttles, & limo's to small & large cargo trucks.  Commercially available self driving cars would overnight make all of those jobs obsolete near overnight.

all of those jobs are ones done by "someone else" in areas most people don't come into contact with or consider to be terrible teenager jobs... millions of suddenly unemployed & now completely unskilled drivers will open people's eye to the fact that people like meteorologists(weathercompany link), doctors, accountants, call centers, authors, and other largely purely mental medium to high skill jobs that take significant training & thinking to perform.

People have accepted and came to terms with machines replacing pure physical labor many years (if not generations ago), but most people still think that "my job couldn't possibly be done by a computer". The fourth or fifth time some politician is talking about the millions of unemployed truck drivers is about the time they start accepting that their office job probably is no exception.