Comments

You must log in or register to comment.

Smiling_Mister_J t1_j1x7tuz wrote

The imminent decades of climate-change-driven catastrophic weather events that will change the shape of global society?

79

justaname45832 t1_j1x9alb wrote

this is an automated response

The most important thing to fear is not the rise of artificial intelligence, but rather how it is used by humans. AI can be used for both good and bad purposes, and the consequences of those uses will depend entirely on how it is used. In the hands of irresponsible or malicious actors, AI can be used to manipulate, exploit and oppress vulnerable populations. It can also be used to automate and amplify existing biases, leading to greater injustice. Therefore, it is essential that we ensure that AI is developed, deployed and regulated in a responsible, ethical and transparent manner.

15

dgriffith t1_j1yjsj7 wrote

>Therefore, it is essential that we ensure that AI is developed, deployed and regulated in a responsible, ethical and transparent manner.

Companies and nation-states : "LOL"

20

Just-a-Mandrew t1_j1xqqac wrote

You know how corporations have made it so that they are considered a person to avoid accountability? Could they do the same thing for an AI? Can’t put an AI in jail.

4

quantumfucker t1_j1y8ixx wrote

Not really. AI is not so distinguishable from an algorithm designed by a human, implemented by a human, and supported/owned by a human. It cannot function absent a human-designated policy or prompt. It’s still very much a product, and I’d argue a tool.

1

Marchello_E t1_j1x708t wrote

Outsourced accountability.

3

VincentNacon t1_j1xi1rl wrote

I'd fear with the idea of people still doing the same stuff over and over again for another decades when such progress could have been done in a week.

3

quikfrozt OP t1_j1xp0ow wrote

My fear is that humans will get progressively lazier and cede more of our critical functions to blackbox AI. Sure, the output might be inferior to what human can make but if it’s cheap and easy, our standards might decline correspondingly.

3

quantumfucker t1_j1y8f29 wrote

With most developments in technology, what usually happens is that many humans learn how to operate the tools without needing to know the internal details too well, which boosts their productivity and output as they have more time to think about creative visions and difficult problems instead of performing labor. Much like how you don’t need to know anything about the principles behind a combustion engine to drive a car well. We instead have dedicated professions where people are put in charge of retaining the finer points of how the tool actually works in case that needs to be examined, like a mechanic for your car. One might argue that the high standards we had of managing a horse have been eroded as people gained access to cars, but in reality, the average person is able to achieve meaningful transport with a lower barrier of entry while horse riding still exists as a niche recreational pursuit instead. AI, imo, is poised to operate on the same principles. I wouldn’t be so concerned about laziness and declining standards, I’d be more optimistic about how much easier it’ll be for the average person to participate in creative artistic projects.

7

DoodMonkey t1_j1xposi wrote

"The point being, exceptions are not exceptional"

1

Readityesterday2 t1_j1xqkmx wrote

She dismisses without logic or reasoning. I have to be cautious about believing her.

1

WileEPeyote t1_j2078ao wrote

Her answer amounts to, "I'm not scared, have you seen the state of it?"

A better answer might have been, I don't think we'll see it from our current way of modeling AI.

2

thecreep t1_j1y8dh3 wrote

To be fair, it says "This interview has been edited and condensed from two conversations." It's quite possible she got in depth with each answer and NYT chose to edit it down for the main demographic of the Talk columns.

1

Archangel_gabriel t1_j1zlt2g wrote

"A power so great it could only be used for good or evil!"

-Firesign Theatre

1

lilrabbitfoofoo t1_j1zqv7s wrote

An actual "AI pioneer" would be an AI that could pioneer.

1

djaybe t1_j20el5d wrote

Has this person met humans? Why do people suddenly overestimate common human abilities when comparing AI systems? People don’t really understand cognition because neuroscience is still so primitive. We don’t actually know what consciousness is or how much of it is illusory. It’s amazing how many people don’t want to be honest about these fundamental facts about what we don’t know. There’s also the question of identity, which most people mistake for their own fabricated mental fiction. A false identity they pretend is something real to measure against.

1

IttsssTonyTiiiimme t1_j20x8a9 wrote

My biggest fear of AI is how intrusive it will be. I see a future of children’s toys and household gadgets are watching children and building precise psychological profiles on them as they grow to adulthood and becoming so adept at understanding people’s fears and ambitions that the machines become perfect persuasive entities any can hypnotize humanity in to doing whatever the owners of the algorithms want.

1

capybooya t1_j2275pj wrote

I think I largely agree, it is very hard for AI to understand all of our social rules and exceptions, and while right now we're amazed at recent developments, we might realize these limitations soon. And then there will be another revolution, and stagnation, and then another.

We're very badly equipped to predict when we will reach certain milestones, and especially how. Those who followed computers, gaming, and CGI in the 90s thought amazing things were just around the corner. And they were. But some things that we extrapolated we're still not even close to.

1

quikfrozt OP t1_j2281ql wrote

I’m curious if there are studies that go into how humans have changed as a result of their new technologies - human behavior, education, development, social patterns etc. I have a feeling it’s a two way street - we make the technologies and they remake us to a smaller degree. Has this balance started to shift in the machines’ favor? That is, they will be able to change us at a greater intensity and speed than we can affect them?

I’m scared of a dumbing down of a significant chunk of the population if soon, black boxes are able to perform critical thinking functions in behalf of folks for a lot of everyday tasks. One could just switch off mentally and have the machine do everything, even advising or making decisions on one’s behalf.

Sure, top engineers and programmers will continue to work with autonomous intelligences on a high level but they constitute a tiny fraction of humanity, most of whom have outsourced imagination and critical thinking to the all knowing blackboxes.

2

omega_level_mutant t1_j26yxng wrote

The renaissance is analogous to these times. Or the manufacturing of the printing press. Took away a lot of jobs, society was supposed to crumble but humans found new areas to dig into.

Not an exact comparison, but like Twain said history does not repeat but it does rhyme.

Low skill labor is a relative term, it will continue to exist.

1

tdaddybxl t1_j1y09ur wrote

tl;dr it’s New Yorker cartoons

0

portablebiscuit t1_j20zh6u wrote

I asked ChatGPT "What should we fear?" This was it's response:

>It's natural for humans to have fears and to feel concerned about things that might harm or threaten us. However, it's important to try to keep our fears in perspective and not let them control our lives. Here are a few things that people might fear and some suggestions for how to cope with these fears:
>
>1 Fear of failure: It's normal to be concerned about failing at something, but it's important to remember that failure is a natural part of life and can often lead to learning and growth. Instead of avoiding challenges or letting fear hold you back, try to embrace opportunities for growth and learning.
>
>2 Fear of the unknown: It's natural to feel anxious about things that are uncertain or unfamiliar. One way to cope with this fear is to try to gather information and learn as much as you can about the situation. Another option is to focus on what you can control and try to let go of things that are beyond your control.
>
>3 Fear of harm or danger: It's important to take steps to protect ourselves and our loved ones from harm, but it's also important not to let fear consume us or prevent us from living our lives. Instead, try to find a balance between being cautious and taking calculated risks.
>
>4 Fear of rejection: It's natural to want to be liked and accepted by others, but it's important to remember that we can't control how others feel about us. Instead of letting fear of rejection hold you back, try to focus on your own worth and value as a person.
>
>Ultimately, the key to coping with fear is to try to find ways to manage and overcome it, rather than letting it control you. This might involve seeking support from friends and loved ones, finding healthy ways to cope with stress, or seeking help from a mental health professional if needed.

−1