It certainly let's us take on more projects which has increased our profits and salaries
. We just hired two more people this week because of the increased work.
Non-paranoid AI thread
- Thread starter UnCytely
- Start date
It certainly let's us take on more projects which has increased our profits and salaries
. We just hired two more people this week because of the increased work.
I know there is already a thread about AI in the Cave, but it is filled with paranoia about SkyNet. I also don't think a reasonable discussion about ChatGPT and AI needs to be political.
I am not worried about "evil AI" taking over, AKA "SkyNet". I am worried about us becoming dependent on bug-riddled AI, which seems likely. More and more people are growing dependent on things like ChatGPT without realizing the limitations and that it can be just plain wrong a lot.
Robot takeover? Not quite. Here’s what AI doomsday would look like
Experts say the fallout from powerful AI will be less a nuclear bomb and more a creeping deterioration of societywww.theguardian.com
My ChatGPT says that post is uncytely
I’m always curious how some read this with comfort, some with concern, and their ages.
It applies even more to some other topics.
With how humanity seems to have a need to destroy itself and how humans tie their worth to work it's impossible for me to not see the potential disaster.
That's before any hiccups to the technology are mixed in.
Bolded is essentially the same argument, just different points. Is the AI is so good when does it take over? Compared to, if the AI enticing enough how far can we go without becoming to dependent, at which point we're screwed so it took over?I know there is already a thread about AI in the Cave, but it is filled with paranoia about SkyNet. I also don't think a reasonable discussion about ChatGPT and AI needs to be political.
I am not worried about "evil AI" taking over, AKA "SkyNet". I am worried about us becoming dependent on bug-riddled AI, which seems likely. More and more people are growing dependent on things like ChatGPT without realizing the limitations and that it can be just plain wrong a lot.
Robot takeover? Not quite. Here’s what AI doomsday would look like
Experts say the fallout from powerful AI will be less a nuclear bomb and more a creeping deterioration of society
At the end of the day regulation will need to be strong, but it will be tough. How do we handle misinformation at a greater scale than we already fail to? How do we handle the economy as more and more jobs are swept up? How do we encourage leaning when more things are available by something simpler than a quick google?
I am fine using it for things like "write me fifteen headings for this report about X that use Y in the title." Things that aren't depending on a lot of facts and figures, or nuance, but rather just aggregating options.
I think it has a lot of uses and 99% of the AIs out there aren't nearly as capable as being stated. They're more of dynamic, search engine 2.0s or grammar and writing learning tools.
With that said... It's now being denied, so who knows if it actually happened, but there was a report last week that was alarming if true...
With that said... It's now being denied, so who knows if it actually happened, but there was a report last week that was alarming if true...
And the general population will lose more critical thinking skills. Making them more vulnerable to deep fakes and people using these tools to scam them and worse.
Reminds me of the book I, Robot.
What safeguard logic can you program into a computer that will prevent the computer determining humans are the greatest threat to humanity?
I'm of the opinion that quantum computers are much like fusion reactors - mind blowing in theory but will be mired in "development" for my entire life. Maybe someday they're viable, but I've lived long enough to doubt that that day will be anytime soon. AI is kinda in the same boat. It's here and it's real, but AI has its limits. As a software developer, I'm not buying into the current AI hype/scare.
Or susceptible to never believing AI is wrong. I don't see either scenario as good.
Garbage in. Garbage out. But more believable.
www.businessinsider.com
AI hypocrisy: OpenAI, Google and Anthropic won't let their data be used to train other AI models, but they use everyone else's content
Generative AI models are trained on a lot of online content. The companies behind these powerful models never asked for permission to do that.
www.businessinsider.com
People are using AI to train AI.
"Using AI-generated data to train AI could introduce further errors into already error-prone models. Large language models regularly present false information as fact. If they generate incorrect output that is itself used to train other AI models, the errors can be absorbed by those models and amplified over time, making it more and more difficult to work out their origins"
www.technologyreview.com
I am not worried about "evil AI" nuking civilization. I am worried about a sophisticated but bug-ridden AI turning all of the traffic lights in a city green all at once (for example), but only for a minute or two, and because it was so transient it becomes difficult to dive into the mountains of data that trained the AI to find out why it happened.
"Using AI-generated data to train AI could introduce further errors into already error-prone models. Large language models regularly present false information as fact. If they generate incorrect output that is itself used to train other AI models, the errors can be absorbed by those models and amplified over time, making it more and more difficult to work out their origins"
The people paid to train AI are outsourcing their work… to AI
It’s a practice that could introduce further errors into already error-prone models.
www.technologyreview.com
I am not worried about "evil AI" nuking civilization. I am worried about a sophisticated but bug-ridden AI turning all of the traffic lights in a city green all at once (for example), but only for a minute or two, and because it was so transient it becomes difficult to dive into the mountains of data that trained the AI to find out why it happened.
I guess I don't understand how it's ok for someone to pull up ChatGPT and type in... "write a blog about this topic that focuses on area 1, area 2, and area 3 as it relates to my profession" and then post that blog with their name on it. I understand that the output is pretty amazing, but putting your name on content that you didn't generate is fukked up.
- Mar 27, 2006
- 40,829
- 28,324
- 113
Agreed. That doesn't seem like something you can truly say that you wrote.
But I had to write a job description for a new position that I'm hiring for, and chatGPT gave me a pretty awesome framework to model it on. I made quite a few changes and tweaks to fit exactly what I was looking for, but I have to say that it did a pretty remarkable job.
That’s exactly how I’ve been using it. It does maybe 80% of the work, but I always end up rephrasing a few things and doing a final edit before sending anything off.
Exactly, I think it'd be perfect for things like that. I've never seen a job description with a byline.