Your thoughts on AI. No arguments, just opinions.

[bherman]

Great topic - what occured to me is that just a few short years ago, most of what we read about the "threat" to jobs and such was automation (ie, robots). What had been overlooked in that angle was the brains behind the automation - and that not all of what was going to be replaced was physical in nature.

 

[mikewalker]

The question to ask is:
Is there a class issue in AI?
Does one section of society own, control and benefit from AI at the expense of the majority of the population?

the amount of $$$$$$$$$$$$ required for the newest AI server farms now under construction (and the electricity to power them!) definitely yields the "yes" answer for that one.

 

[Dr. Love]

The potential of the technology simultaneously excites me and terrifies me.

 

[thewildest]

The core problem with AI, is “A”…. From now, it is limited to hyper fast access to extended amounts of data, following code written by people, so inherently limited to what humans can observe, document and code. Yes, we can “iterate” faster, but through a limited set of data, mostly restricted to our individual and segmented perception. While many consider that already as a threat, there really are not supra-human capabilities, otherwise we would have already answers for some of our basic unsolvable equations (cure to cancer, how to solve global warming, how to link astronomy with quantum mechanics, etc.). None of that, today, can be answered by AI.

So… let’s chill…. Until AI will observe, select data points and program itself. The day an entity would develop its own logical pathways to solve problems (what a basic cell does), it could quickly gain a primitive way of consciousness. From that point forward, we will be not longer needed as the “carriers” of consciousness and therefore we’ll become obsolete nuisances slowing down its advance.

 

[ahbradot]

I liked each and every post just to make a point that I love that we are taking about this very important subject amongst us carbon lifeforms.
No I am not a bot.
I just support humanity.
I do not care for this AI uncertainty.... it make me feel vulnerable.

 

[TeaTime]

I'm not concerned that the machines will take over - I'm concerned what those who already wield extreme power will do with this powerful tool. We are going to see job displacement on a scale beyond imagination. Many in America value dollars before social stability or human rights, and that's not going to change. Those in power are not going to care for those who are left desperate and hopeless. They will instead blame the minimum wage guy who just lost his job, blame the whole situation on him, because that's their playbook.

 

[Carl Hillman]

I have to laugh at the idea of AI destroying humanity. Your thoughts?

I'll still make the occasional "Skynet" joke, but after reading Superintelligence: Paths, Dangers, Strategies I don't laugh much anymore.

 

[CryingBass]

I think we all need to take the long-time patented approach to dealing with AI, developed by the guru of all things such, Jerry Garcia..

Wait - what was I saying??

 

[james condino]

Would AI chose tort?

 

[dbsfgyd1]

I've played around with AI. Mostly just for realizing art projects. I see a lot of positives, but also some potential negatives. I've been a fan of cyberpunk novels in the past and I've always thought that they were merely entertaining. Some of the dystopia presented in the stories seems to be becoming real now, but then again, I remember the Y2K paranoia and I have to laugh at the idea of AI destroying humanity. Your thoughts?

Who knows? I have no preconceived bias nor do I fear it. As far as music goes, I've known I'm a dinosaur heading to the tar pits long ago. However, I would love to play around with it.

Years ago we had a Yamaha with pre programmed Musical styles. You played the chords and it would figure out the rest of the the rhythm section. WE too much joy in throwing all sorts of abrupt and weird chord and odd timings to see how it would resolve it by the first beat of the next measure. It was hilarious.

 

[fdeck]

What heartens me is that people are actually thinking about the working class, social conditions, etc.

But the "post truth era" predates AI by a few years, and people could already by swayed to embrace outrageous propaganda that they knew to be false, in 1939 if not earlier.

I'm just a little bit skeptical, because a lot of the "experts" seem to be taking a Pascal's Wager approach of imagining the most extreme consequences -- an intelligence that's self-assembling, omniscient, and omnipotent -- with appropriate dread. The problem with such discussions is that the outcome can be pretty much anything you can dream up, if there are no limitations.

It's interesting to notice that there are still some things AI can't do. It can't discern the truth. Well, as it turns out, neither can humans. 3+ millennia of philosophers still haven't reached an agreement on what "truth" consists of, or how to reliably know if something is true. We don't know if it can make things that are reliable (related to truth). So far humans have been able to automate some jobs, but have not figured out how to remove the human who can intervene manually when the automation fails, which is often.

Technology has changed education in the past. I was the first student at my college to submit a word processed term paper. I think it gained me at least 1/2 point on my overall GPA because it dramatically changed how I wrote papers. In fact I deliberately signed up for courses where the grading was dominated by written work.

 

[Jogobass]

We can't make general AI because we (science) don't know what consciousness is. However, if consciousness emerges in AI it will be able to tell us what it is.

 

[bobdabilder]

You have absolutely nailed it.

It's terrifying inasmuch as objective truth has been replaced by the Tower of Babble that is social media, so 'real' and 'authentic' are about to be stricken from the modern world by . . . . software.

 

[MarshallAC50Stack]

Things would look a lot better if we’d learned how to deal with the basics before this question came up.

 

[ebo]

There will undoubtably be benefits to humanity as a whole. Those will mask the biggest benefits that’ll go to big business, government concerns, and big power brokers. Will AI be used to create the ultimate weapon? Are people already working on it? AI is a tool and humans misuse their tools. Ask my basses, they’ll agree.

 

[_Some Dude]

IMO, AI will be disruptive, but it will also be liberating.

If you consider previous rounds of automation, I don't see people lamenting that they are no longer required to dig ditches. Nor do I see people complaining that they are no longer required to dig potatoes out of the ground. The backhoe and the combine have liberated thousands upon thousands of people to do more productive labour.

Yes, some people will be out of work because their job will be made redundant (elevator operator, punch card operator, telephone switchboard operator), but society will move on and humans will move on to doing different things.

 

[bobdabilder]

Imagine a world where you can't find the truth, historical accounts, news of robberies or general crime, actual stock market conditions, etc. You only receive the information it wants you to have. Only points you to certain stores or restaurants. The internet is already biased this way by people. Imagine a place where a biased ai is controlling most of the information we get from the internet.

 

[Robot.G]

Would AI chose tort?

Ask it!

 

[MarshallAC50Stack]

I'm heading to the practice room...

 

[Carl Hillman]

We can't make general AI because we (science) don't know what consciousness is.

I'd add, we can't instill human values in AI because we can't agree on exactly what they are (see: the Trolley Problem or any other classic ethical dilemma).