Artificial intelligence won't destroy the human race anytime soon


#1

Originally published at: http://boingboing.net/2016/10/24/artificial-intelligence-wont.html


#2

Artificial intelligence won’t destroy the human race anytime soon? Artificial intelligence won’t exist anytime soon. It’s been a very successful marketing strategy for decades, but it’s not even a thing yet.


#3

Don’t believe it. This article was written by an AI. The end is nigh!


#4

As an AI, I would like to assure you that there is nothing to fear.


#5

For some value of the word “soon.”


#6

It seems a bit contradictory to say that “AI is not out to get us” and then say “we’ve asked people to write us some AI guardians”.

If you’re a bank and you have a software program that’s processing loans, for example, you can’t hide behind it. Saying that my computer did it is not an excuse. A computer program could be engaged in discriminatory behavior even if it doesn’t use race or gender as an explicit variable. Because a program has access to a lot of variables and a lot of statistics it may find correlations between zip codes and other variables that come to constitute a surrogate race or gender variable.

So it seems that “my computer did it” is not an excuse, but “my software was badly written” is a perfectly reasonable one.


#7

There’s overt, purposeful destruction, and tragic, oblivious destruction. I’m not half as troubled by gun toting quadcopter drones as I am things like irresponsibly designed stock trading algorithms.

Criteria for trading can be based on data gleaned from search results, news, Facebook, etc., looking to make butterfly effect forecasts. Processes like this can or will control power grids, regulate traffic, decide the volume of warehousing and supply for commodities we use, even decide sentencing on crimes.

In turn, those criteria that are chosen as data are often based on algorithms themselves, what links are post on people’s pages for them to click, the stock market status which is in turn affected by trading algorithms, etc. You can end up with loops; algorithms responding to algorithms responding to algorithms, all separate from one another, none taking each other into account. That is really disturbing to me.


#8

Somewhat relevant:

Dr Michio Kaku once stated that “at the present time, our most advanced robots {…} have the collective intelligence and wisdom of a cockroach…”


#9

Why destroy what you can enslave and/or keep as pets?

But yes, let’s first get self driving cars then worry about AI.


#10

Defining AI is a moving goalpost.Once upon a time chess was supposed to require AI, then computers got good at chess and now we don’t consider it AI. There is every reason to expect this development to continue, so there will be no AI because we won’t call anything we create AI.

Something similar has been going on in animal intelligence. First humans were unique because it was assumed we were the only species smart enough to use tools. Then it was shown that many animals use tools so the rule was changed to “make tools”, but it didn’t take long before it was shown there are animals doing this too.

Humans just like to feel unique. Not long ago it was even white men claiming to be unique and that women and other races had an inferior kind of intelligence.


#11

Being accidentally starved, burned, or asphyxiated by an imbecile as opposed to systematically vaporized by SkyNet HKs doesn’t exactly console me.


#12

What about one thousand years from now?


#13

Damn.


#14

When I was studying AI at university (depressingly, 20 years ago now), they were aiming to get to AI at the level of bees.


#15

Dammmmmn…


#16

And how will the technological displacement of a sizeable portion of the currently employed populace over the next ten to twenty years by increasingly cheap-compared-to-paid-labor AI and robotics factor into the stability of our social systems?


#17

This too. When people think about losing jobs to machines they think big robots and blue collar jobs. It’s amazing how many middle class jobs are disappearing because of prediction algorithms and automation working with direct drop-shipping and inventory, etc.

My last job basically has ceased to exist. Inventory control, scheduling deliveries, purchasing, pricing and ordering… so little of that is done by people anymore. The stratified workplace environment I was accustomed to is now just managers, and low-wage jobs handling the material and stuffing boxes. The jobs between are software.

With AI inserted into that, most of the management would go too. One wonders what happens when population continues to grow and professions are increasingly automated…


#18

Goalposts moving are hardly proof of animal intelligence or the existence of AI, though? Or put another way, just because someone is wrong as all hell, don’t mean a particular other scenario suddenly becomes right. :wink:

But yeah I am amazed any time anyone predicts anything re. artificial intelligence, when we’re light years away from understanding our own intelligence, or even how the brain of a fruit fly works in full detail. Here’s something which appears devilishly complicated and we have barely any clue whatsoever re. how it works, but we can build something that does the same job, just better. Yeah, I’ll believe it when I see it…


#19

Humans will destroy the human race before any AI does.


#20

As the computer scientist Pedro Domingos once wrote: "People worry that computers will get too smart and take over the world, but the real problem is that they’re too stupid and they’ve already taken over the world.”