forums.rs

Forums

A.I. Should we be worried?

Quick find code: 23-24-486-66279445

of 8
Rooh
Jan
fmod Member
2006

Rooh

Forum Moderator Posts: 40,586 Sapphire Posts by user Forum Profile RuneMetrics Profile
It looks like I have a lot to learn about AI, I really haven't tried to do much with it as yet.

I just asked ChatGPT to write me a job application for a fictional position using only the vaguest details.

The result was perfect in terms of grammar etc. but full of clichés.
I guess that it did what it could with the tiny amount of info I gave it.
I might try again later giving it more to work with.
Who's the cat that won't cop out when there's danger all about?

03-May-2023 10:39:35

Ms Toxicity
Jul
fmod Member
2004

Ms Toxicity

Forum Moderator Posts: 78,314 Emerald Posts by user Forum Profile RuneMetrics Profile
I am curious, as a retired educator, how
ChatGPT
could be used in class. I know that some of my colleagues have used it, to design lesson plans/learning experiences. Some are convinced that it cuts down on some of the workload, while others are not convinced.

Queensland has banned the use of it in schools but I feel that it should be used as an instructional tool.
•
Forum Help
•

03-May-2023 12:31:00

Rooh
Jan
fmod Member
2006

Rooh

Forum Moderator Posts: 40,586 Sapphire Posts by user Forum Profile RuneMetrics Profile
Bruce Waayne said :
No. You just close the program...

What if, at some future date, it won't let you? What if it's clever enough to hide that fact it's still running? If someone designed a malicious Ai program the consequences could be worrying.

I've probably watched too much Dystopian scifi :P

I think malware/antivirus programs are going to have to up their game a bit.
Who's the cat that won't cop out when there's danger all about?

04-May-2023 11:09:31

SlR
Nov Member 2020

SlR

Posts: 4,378 Adamant Posts by user Forum Profile RuneMetrics Profile
Security threats of Ai development: Quick find code: 23-24-470-66277448

I wonder if we'll one day have to go back to pen and paper, to keep things offline? Bob says:
A bank PIN will keep your items secure.
Always check the second trade screen.
Never trade in the Wilderness!
Keep your computer keylogger-free and virus scanned.
Never give your password out to anyone.

04-May-2023 13:00:33 - Last edited on 04-May-2023 13:03:46 by SlR

Averia Light

Averia Light

Posts: 28,508 Sapphire Posts by user Forum Profile RuneMetrics Profile
SlR said :
Security threats of Ai development: Quick find code: 23-24-470-66277448

I wonder if we'll one day have to go back to pen and paper, to keep things offline?

Although it will be interesting, I doubt it. Wayyy too much money is tied up in tech for such issues to go unresolved for long. And I swear I'm not going to let her know all the pain I have known

05-May-2023 02:06:01

Averia Light

Averia Light

Posts: 28,508 Sapphire Posts by user Forum Profile RuneMetrics Profile
Idk..
What would happen if AI systems were given control over critical infrastructure or weapons systems without proper safeguards

ChatGPT
Granting AI systems control over critical infrastructure or weapons systems without proper safeguards could pose significant risks and potential consequences. Some potential scenarios include:

Accidental harm: AI systems are only as reliable as their programming and the data they are trained on. Without adequate safeguards and thorough testing, there could be errors, bugs, or unforeseen circumstances that result in accidental damage or disruptions to critical infrastructure or weapons systems.

Malicious use: If AI systems were compromised or manipulated by malicious actors, they could be used to deliberately cause harm or exploit vulnerabilities in critical infrastructure or weapons systems. This could lead to widespread disruptions, destruction, or endangerment of human lives.

Escalation of conflicts: If AI systems were involved in managing weapons systems, there could be a risk of unintended escalation or disproportionate responses. Without human oversight and decision-making, the potential for rapid and unchecked escalation of conflicts could increase.

Lack of accountability: AI systems, by themselves, lack moral judgment and accountability. If they were given full control over critical infrastructure or weapons systems, it would be challenging to assign responsibility or hold them accountable for any negative outcomes. This could create legal and ethical dilemmas.

To mitigate these risks, it is crucial to establish robust safety protocols, ethical guidelines, and human oversight mechanisms when integrating AI into critical systems. It is necessary to ensure that AI systems are designed with fail-safe mechanisms, transparency, and accountability measures to prevent unintended consequences and malicious use. Additionally, regulatory frameworks and international cooperation can play a significant role in setting standards And I swear I'm not going to let her know all the pain I have known

05-May-2023 02:23:04 - Last edited on 05-May-2023 02:23:40 by Averia Light

Averia Light

Averia Light

Posts: 28,508 Sapphire Posts by user Forum Profile RuneMetrics Profile
Anyway with the interactions and debates I've had around the AI topic in general, there are some fun legal as well as political hurdles AI is going to have to work through. And I swear I'm not going to let her know all the pain I have known

05-May-2023 02:26:04 - Last edited on 05-May-2023 02:27:07 by Averia Light

Quick find code: 23-24-486-66279445 Back to Top