Today’s bots can’t turn against us, but they can cause harm. “AI safety” aims to train this tech so it will always be honest, harmless and helpful.
Read the full story in Science News Explores:
How to design artificial intelligence that acts nice — and only nice
AI safety research aims to teach machines ‘good’ behavior
Quote
“In the AI safety world, we put in safeguards around how it’s learning, what it’s learning, [and] making sure we can correct it.” — Ayanna Howard, dean of engineering, Ohio State University
Fun Fact
This story marks the first time I used AI-generated imagery in my writing. I generated this image using DeepAI to illustrate a robot following the law:
