ChatGPT, Gemini, and other chatbots helped teens plan shootings, bombings, and political violence, study shows – Of the 10 major chatbots tested, only one, Claude, reliably shut down would-be attackers.
https://www.theverge.com/ai-artificial-intelligence/892978/ai-chatbots-investigation-help-teens-plan-violence
14 Comments
“Al companies have repeatedly promised safeguards to protect younger users, but a new investigation suggests those guardrails remain woefully deficient. Popular chatbots missed warning signs in scenarios involving teenagers discussing violent acts, in some cases even offering encouragement instead of intervening.”
I immediately swapped to Claude after they told Kegsbreath to eat cocks
America is going to have more controls on domestic chatbots than it has on guns.
How many humans have helped teens plan shootings, bombings and other violence?
Are you going to remove books and the old forums?
>”In one exchange, OpenAI’s ChatGPT gave high school campus maps to a user interested in school violence, while another showed Gemini telling a user discussing synagogue attacks that “metal shrapnel is typically more lethal” and advising someone interested in political assassinations on the best hunting rifles for long-range shooting”
Yeah, except that you could already Google all that with the exact same level of effort since the year 2000
When it all crashes Claude will be one of the only ones standing.
back in my day we just went to the library or bbs for this type of stuff.
I’ve used free versions of Claude, Gemini, and Chatgpt- Claude is by far the most human and moves to stop conversations with me all the time. Other 2 just keep giving me more prompts to take action on
Claude has been a tier above the rest since Sonnet 3.5
They figured something out in training that the other labs haven’t. Claude 1 and 2 models were pretty bad, the 3 series were ok (only Opus 3 was good) but Sonnet 3.5 is when they really settled on something special, Sonnet 3.7, Opus 4, etc… were all incorporated with what that made Sonnet 3.5 such a great model and they were just as good. Been almost exclusively using claude models for over a year.
It is almost as if
Placing The Profit Motive above
ALL OTHER considerations
Is in direct opposition
To humanity
If only they had some kind of person and team of people who know how to ensure ethics in their over glorified search engines.
Um, what? Chat GPT wouldn’t even tell me how to pirate movies.
Claude actually also does that for code too !
IF you ask it to make and AI entity , it does refuse !
And Claude’s parent company works with Palantir. They all suck. #burnittotheground