Researchers use ASCII art to elicit harmful responses from 5 major AI chatbots
LLMs are trained to block harmful responses. Old-school images can override those rules.
Researchers use ASCII art to elicit harmful responses from 5 major AI chatbots
LLMs are trained to block harmful responses. Old-school images can override those rules.
@arstechnica No doubt they'll try to make filters for this too, but until such time as the filter system becomes even far more complex than the LLMs themselves, there will always *ALWAYS* be some way people find around whatever filters they try to write.
Until people stop treating LLMs as if they were actually AI, this is just going to keep going on and on and on.
076萌SNS is a social network, courtesy of 076. It runs on GNU social, version 2.0.2-beta0, available under the GNU Affero General Public License.
All 076萌SNS content and data are available under the Creative Commons Attribution 3.0 license.