Jump to content

Artificial Intelligence: Are You Concerned?


cddolphin

AI  

24 members have voted

  1. 1. What are your thoughts on AI?

    • AI could very well turn into an existential threat if humans aren't careful
    • AI could lead to big problems, but it's nothing we can't handle as a species
    • AI fear is overblown, any problems should be minimal and easily overcome
    • Fears about AI? People were worried about Y2K too. It will only bring good, any problems will be human error


Recommended Posts

6 minutes ago, theJ said:

Probably could say a lot of theoretical stuff, but i think the best answer to this question is: because history says it will happen.  

Machines have been replacing human labor for centuries.  We always find work for those displaced by them.  We also keep expanding the population.

Those people will find something to do.

Yeah basically this. Unemployment in the West is reasonably low despite all the technological advances we've achieved since the industrial revolution. People find work in emerging fields.

Link to comment
Share on other sites

On 4/6/2018 at 10:55 AM, texans_uk said:

I'm hoping we are building reliable fail safes!

As a pseudo-developer, I'm not too concerned.  AI will rely on code and all code is flawed (or can be exploited) in some way.  So even if there isn't a backdoor explicitly coded in, doesn't mean there isn't one.

AI can be a threat and might cause damage, but I think it's damage that can be handled by the human race which is why I picked option #2...

Link to comment
Share on other sites

7 minutes ago, theJ said:

Probably could say a lot of theoretical stuff, but i think the best answer to this question is: because history says it will happen.  

Machines have been replacing human labor for centuries.  We always find work for those displaced by them.  We also keep expanding the population.

Those people will find something to do.

sure, but we've never had intelligence that was smarter than humans before.

Link to comment
Share on other sites

1 minute ago, TheVillain112 said:

As a pseudo-developer, I'm not too concerned.  AI will rely on code and all code is flawed (or can be exploited) in some way.  So even if there isn't a backdoor explicitly coded in, doesn't mean there isn't one.

AI can be a threat and might cause damage, but I think it's damage that can be handled by the human race which is why I picked option #2...

or it means there is a failsafe but it doesnt work because it can be exploited by the vastly superior AI.

Link to comment
Share on other sites

20 minutes ago, mistakey said:

sure, but we've never had intelligence that was smarter than humans before.

And we still don't.  

If the argument is what will happen when we have robots that can think like a human, than sure.  We can start to worry.  But that is not looking likely in our lifetime without a huge breakthrough.

Link to comment
Share on other sites

23 minutes ago, mistakey said:

or it means there is a failsafe but it doesnt work because it can be exploited by the vastly superior AI.

I've never heard of a machine that can program itself.  Learn, sure.  But not re-program its base code.

Link to comment
Share on other sites

I think people watch movies like I Robot and assume we have working prototypes of that.  But that's all fiction people.  That kind of intelligence doesn't exist, anywhere.  Not even close.

We spend millions of dollars to design and program one machine to replace the task of a human that is being paid $11/hr.  And the product we get is a rigid machine that can't deviate from its original purpose even one inch.  If we had the kind of intelligence we're talking about being dangerous to us, those machines could replace $30/hr humans doing engineering and they'd already be everywhere.

Link to comment
Share on other sites

1 hour ago, theJ said:

I've never heard of a machine that can program itself.  Learn, sure.  But not re-program its base code.

thats the worry of General Artificial Intelligence, right?

we created AI that beat kasparov in chess in 95, and just finished go.  i think you're underestimating how soon it could come, especially since this is the new race to a nuclear weapon for every large nation state.  

Link to comment
Share on other sites

furthermore, say someone writes an AI program that is supposed to maximize human pleasure.

Whats to say that AI doesnt realize after multiple iterative failures after its out of the sandbox that the best way to maximize human pleasure is to capture all humans, openup the skull and then send electrical signals to the pleasure center of the brain?

Link to comment
Share on other sites

On 4/6/2018 at 8:39 PM, theJ said:

That's a long way off too.  We've largely picked off the easy to automate working class jobs.  There aren't as many slam dunks as there used to be.  Humans are extremely flexible (if imperfect) workers.  Much of what we're trying to automate today requires a lot of sophisticated AI, which doesn't usually have a business case (i.e. doesn't pay for itself).

If we find a way to lower the cost of automation dramatically, then you start to worry.  But today, the fear of robots eliminating jobs should be pretty minimal.

There's also the flip side to the argument, that these things just make us wealthier as a whole and moves jobs away from manufacturing and into the service industry. 

Human labor still has a lot of value.

AI is entering the service industry at a very fast rate as well though. Look at McDonalds. They are replacing their cashiers with automated ordering systems.. Hotels, nowadays you can book a room, select your room, and receive your key on your phone without ever even seeing or speaking with a rep.

Link to comment
Share on other sites

On 4/6/2018 at 8:39 PM, theJ said:

That's a long way off too.  We've largely picked off the easy to automate working class jobs.  There aren't as many slam dunks as there used to be.  Humans are extremely flexible (if imperfect) workers.  Much of what we're trying to automate today requires a lot of sophisticated AI, which doesn't usually have a business case (i.e. doesn't pay for itself).

Next two major targets for AI breakthroughs in terms of severely disrupting the labor pool: law and medicine.

Doesn't have to be "blue collar."

Link to comment
Share on other sites

6 hours ago, cddolphin said:

Self-driving cars will likely eliminate 10% of jobs within a decade or two. There'll be some pain while the workforce is re-allocated but I just the free market will find create work for them.

Current size of the labor force in the US is ~160M. 16M people suddenly having no job and no job prospects is a major problem.

There's also the 2nd and 3rd order effects: insurance salespeople (yes, don't like them, but they are jobs), repair shops, people in aftermarket parts supply chain, gas station attendants, truck stop workers, people working a dealerships ... those are just off the top of my head.

want a self-driving car. At the same time, I'm not ignoring that this will be incredibly disruptive to the labor market (and potentially the societal fabric).

Link to comment
Share on other sites

Archived

This topic is now archived and is closed to further replies.

×
×
  • Create New...