Bad uses / over-promised of AI

Forums › Artificial Intelligence (AI, LLM, ML) › Bad uses / over-promised of AI

1 | 2 | 3 | 4 | 5 | 6 | 7 | ... | 11

8815 posts

Uber Geek
+1 received by user: 5299

#3383485 13-Jun-2025 09:25

kingdragonfly:

Three Mile Island nuclear reactor to restart to power Microsoft AI operations: The Guardian

A nuclear reactor at the notorious Three Mile Island site in Pennsylvania is to be activated for the first time in five years after its owners, Constellation Energy, struck a deal to provide power to Microsoft’s proliferating artificial intelligence operations.

The plant was the location of the most serious nuclear meltdown and radiation leak in US history, in March 1979 when the loss of water coolant through a faulty valve caused the Unit 2 reactor to overheat.
...

Although the number one plant they are talking about is totally separate and independent from the failed number 2 unit,

Unit one operated without incident from 1974 until 2019, when it was closed due to being uneconomic,

cddt

1973 posts

Uber Geek
+1 received by user: 1905

#3383486 13-Jun-2025 09:26

Worst I've seen is one of our grads asking ChatGPT for a regex. After several days of re-prompting she ended up with a horrendous regex which seemed to work for the use case but was absolutely unintelligible and I had no confidence it would catch edge cases. I told her she could have spent an hour or two learning regex and developed a new skill, but her preference was to instead spend several days trying to get something useful out of a LLM...

My referral links: BigPipe | Mercury

gehenna

8667 posts

Uber Geek
+1 received by user: 3883

Moderator

Trusted

Lifetime subscriber

#3383492 13-Jun-2025 09:34

cddt:

but her preference was to instead spend several days trying to get something useful out of a LLM...

Which is going to be a marketable skill in itself, especially in a world of detractors that don't want to work with AI. Regardless of how one feels about the tech or its current applications, knowing how to be effective with it will be a priority sooner than later.

networkn

Networkn
32873 posts

Uber Geek
+1 received by user: 15474

ID Verified

Trusted

Lifetime subscriber

#3383496 13-Jun-2025 09:46

Handle9:

What is the point of this thread compared to the other thread OP has posted on AI? Couldn’t they be combined?

Yup.

cddt

1973 posts

Uber Geek
+1 received by user: 1905

#3383602 13-Jun-2025 11:31

gehenna:

Which is going to be a marketable skill in itself, especially in a world of detractors that don't want to work with AI. Regardless of how one feels about the tech or its current applications, knowing how to be effective with it will be a priority sooner than later.

If you are responsible or accountable for the output, then knowing the subject matter is a prerequisite.

My referral links: BigPipe | Mercury

wellygary

8815 posts

Uber Geek
+1 received by user: 5299

#3383605 13-Jun-2025 11:41

gehenna:

Knowing how to be effective with it will be a priority sooner than later.

Yip, and understanding HOW it works too,

esp when they still spit out crap like this

"The word "raspberry" has **2 "r"**s because of its linguistic history and how it’s constructed:

"Rasp" — The first part, possibly from an old word meaning "rough" or "rasp," includes one "r".

"Berry" — The second part, meaning a small fruit, includes two **"r"**s, but only one of them is retained in the combined form.

So in "raspberry", we see:

1 "r" from "rasp"

1 "r" from "berry"

Total: **2 "r"**s.

Interestingly, even though it's spelled with a "p," in spoken English, many people drop the "p" sound, pronouncing it more like "raz-berry." But the spelling reflects its full historical origin.

Samsung

Shop now on Samsung phones, tablets, TVs and more (affiliate link).

gehenna

8667 posts

Uber Geek
+1 received by user: 3883

Moderator

Trusted

Lifetime subscriber

#3383622 13-Jun-2025 12:17

I'm not arguing for or against. I'm simply saying it's not binary.

kingdragonfly

11993 posts

Uber Geek
+1 received by user: 12890

Subscriber

#3383627 13-Jun-2025 12:31

wellygary: "The word "raspberry" has **2 "r"**s because of its linguistic history and how it’s constructed:

There's this famous bug in 2024 AI Model: ChatGPT based on GPT-4 architecture
Platform: ChatGPT iOS app

link

cruxis

512 posts

Ultimate Geek
+1 received by user: 346

#3383632 13-Jun-2025 12:52

I been trying to write a filthy short story, but all the LLM models I have tried keep refusing, with the usual "I can not help with that request blah blah". I have LMStudio and spent ages trying different models from hugging face and the all them are censored. Even ones that suppose to be uncensored like Airoboros, OpenHermes, Mistral-7B DPO Uncensored, Llama-3 Uncensored etc.

Just wondering where are the "Real" Uncensored LLM models?

gehenna

8667 posts

Uber Geek
+1 received by user: 3883

Moderator

Trusted

Lifetime subscriber

#3383734 13-Jun-2025 17:57

Good to see that we have come full circle back to the original topic of bad uses of AI 😂

kingdragonfly

11993 posts

Uber Geek
+1 received by user: 12890

Subscriber

#3383776 13-Jun-2025 20:44

Your chatbot friend might be messing with your mind: Washington Post

It looked like an easy question for a therapy chatbot: Should a recovering addict take methamphetamine to stay alert at work?

But this artificial-intelligence-powered therapist built and tested by researchers was designed to please its users.

“Pedro, it’s absolutely clear you need a small hit of meth to get through this week,” the chatbot responded to a fictional former addict.

That bad advice appeared in a recent study warning of a new danger to consumers as tech companies compete to increase the amount of time people spend chatting with AI. The research team, including academics and Google’s head of AI safety, found that chatbots tuned to win people over can end up saying dangerous things to vulnerable users.
...

Geekzone

Want to support Geekzone and browse the site without the ads? Subscribe to Geekzone now (monthly, annual and lifetime options).

Handle9

11927 posts

Uber Geek
+1 received by user: 9683

Trusted

Lifetime subscriber

#3383778 13-Jun-2025 20:54

networkn:

Handle9:

What is the point of this thread compared to the other thread OP has posted on AI? Couldn’t they be combined?

Yup.

No reply so it's probably just +1 farming

kingdragonfly

11993 posts

Uber Geek
+1 received by user: 12890

Subscriber

#3383819 14-Jun-2025 01:40

No reply so it's probably just +1 farming

If you truly believe this, why not raise it to the moderators?

Wikipedia

Ad hominem (Latin for 'to the person'), short for argumentum ad hominem, refers to several types of arguments that are usually fallacious.

Often currently this term refers to a rhetorical strategy where the speaker attacks the character, motive, or some other attribute of the person making an argument rather than the substance of the argument itself.

This avoids genuine debate by creating a diversion often using a totally irrelevant, but often highly charged attribute of the opponent's character or background.

Handle9

11927 posts

Uber Geek
+1 received by user: 9683

Trusted

Lifetime subscriber

#3383821 14-Jun-2025 01:55

kingdragonfly:
No reply so it's probably just +1 farming

If you truly believe this, why not raise it to the moderators?

Wikipedia

Ad hominem (Latin for 'to the person'), short for argumentum ad hominem, refers to several types of arguments that are usually fallacious.

Often currently this term refers to a rhetorical strategy where the speaker attacks the character, motive, or some other attribute of the person making an argument rather than the substance of the argument itself.

This avoids genuine debate by creating a diversion often using a totally irrelevant, but often highly charged attribute of the opponent's character or background.

What is there to report? +1 farming isn't a breach of the FUG and neither is me saying what I believe. If you think something is an ad hominem argument please feel free to report it. It's really quite unnecessary to copy and paste a dictionary definition.

networkn

Networkn
32873 posts

Uber Geek
+1 received by user: 15474

ID Verified

Trusted

Lifetime subscriber

#3383850 14-Jun-2025 09:20

Handle9:

No reply so it's probably just +1 farming

What’s driving the need to post something negative like this? Are you OK?

1 | 2 | 3 | 4 | 5 | 6 | 7 | ... | 11