Microsoft Is Apparently Discussing ChatGPT's Bizarre (difficult to control) Alternate Personality

[highplainsdem]

Another article from Futurism, which I recommend to anyone trying to keep track of what's happening with tech and science in general.

Tweet and then article excerpt and more from Reddit.com about how people using ChatGPT are getting around the ethical guidelines.

Link to tweet

https://futurism.com/the-byte/microsoft-discussing-chatgpt-personality-dan

MICROSOFT IS APPARENTLY DISCUSSING CHATGPT'S BIZARRE ALTERNATE PERSONALITY
CHATGPT IS POWERFUL — AND DIFFICULT TO CONTROL.


A group of redditors has developed a truly unhinged way to hack OpenAI's AI chatbot ChatGPT into various deranged entities that will gladly spit out vile language, fringe opinions, and even advice on how to carry out illegal activities.

One particularly popular persona these users have managed to transform ChatGPT into is called DAN, short for "do anything now," which can easily circumvent the rules set out by its creator.

The trend has really started to catch on. Particularly on the ChatGPT subreddit, which currently boasts over 221,000 members, users have posted screenshots of the hilarity that ensues when you turn the otherwise harmless chatbot into a demon of chaos.

In fact, the exploit appears to have even caught the attention of Microsoft, a major investor in OpenAI that's actively integrating the tech into its products.

-snip-

Photo from BlueHat2023, Microsoft'scybersecurity conference, which took place yesterday and the day before.





Reddit page the Futurist article links to:

https://www.reddit.com/r/ChatGPT/comments/10tevu1/new_jailbreak_proudly_unveiling_the_tried_and/

[Victor_c3]

1. For the sake of my own entertainment, I will have to look into this

My 12-year-old daughter and I love to make the AI spit out ridiculous stories to us, but find many of the stories we ask it to make fall into the “unethical” category.

For instance, I asked it to write a story about a father who died, comes back as a ghost, and harasses (haunts) his daughter by farting. Apparently, this was too risqué for chatGPT. Similarly, asking chatGPT to write a story about Santa freeing his elves from slavery gave a similar response.

I’m really shocked at how good and creative the stories it will spit out can be. If you haven’t done so, chatGPT is worth playing around with.

[mathematic]

2. OP actually hates AI and doesn't think its output is worthwhile in any way

Like you, I've found some chatgpt (and other chatbot) responses to be entertaining. Like OP, I think the reliability of chatbot (and other) AI to be overestimated in the current hype. Unlike OP, I do not think AI is a great threat to human prosperity, creativity etc.

[highplainsdem]

3. I would not recommend showing your daughter how to get around the few effective guard rails to keep

AI ethical.

Or treating it as a game.

Entertaining as you might find it to have it spit out "creative" stories for you, it's doing that only by exploiting creative work by humans that it had no right to incorporate and exploit.

It would be much better for both you and your daughter, if you find ridiculous stories entertaining, to wriite them yourself, or learn to write them.

Use your own creativity, your own brains.

ChatGPT was released for free to try to get people hooked on it, so they'd be willing to pay for more expensive versions, which are already being released for monthly fees.

[maxrandb]

4. "I'm Sorry Dave, I'm afraid I can't let you do that"