Meta unleashes BlenderBot 3 upon the internet, its most competent chat AI to date
, the incident still stands as stark reminder of how quickly an AI can be corrupted after exposure to the internet's potent toxicity and a warning against building bots without sufficiently robust behavioral tethers. On Wednesday, Meta's AI Research division will see if its latest iteration of Blenderbot AI can stand up to the horrors of the interwebs with the public demo release of its 175 billion-parameter Blenderbot 3.
"Researchers can’t possibly predict or simulate every conversational scenario in research settings alone," Meta CEO Mark Zuckerberg wrote in a Wednesday blog post."The AI field is still far from truly intelligent AI systems that can understand, engage, and chat with us like other humans can. In order to build models that are more adaptable to real-world environments, chatbots need to learn from a diverse, wide-ranging perspective with people 'in the wild.
When a user logs an unsatisfactory response from the system— currently hovering around 0.16 percent of all training responses — Meta works the feedback from the user back into the model to avoid it repeating the mistake. The system also employs the Director algorithm which first generates a response using training data, then runs the response through a classifier to check if it fits within a user feedback-defined scale of right and wrong.
France Dernières Nouvelles, France Actualités
Similar News:Vous pouvez également lire des articles d'actualité similaires à celui-ci que nous avons collectés auprès d'autres sources d'information.
AI Ethics Fighting Passionately For Your Legal Right To Be An ExceptionAI Ethics is fighting for your legal right to be an exception when it comes to AI autonomous systems that are based on a no-exceptions Machine Learning approach.
Lire la suite »
AI Facial Recognition Systems Work the Worst for Black WomenAs scientists work to design and test robots and their perceived intelligence, Dr. Joy Buolamwini is dedicated to fighting algorithmic bias in order for AI systems to recognize individuals from marginalized communities.
Lire la suite »
DALL-E Mini: Everything you need to know about the AI art generator ruling the internetWith the launch of AI text-to-image generator DALL-E Mini, the internet has seen hilarious images of cartoon crossovers, warped celebrities and surreal imagery.
Lire la suite »
Ambitious Researchers Want to Use AI to Talk to All AnimalsA group of animal lovers is looking to use machine learning to translate animal 'language' into something humans can understand.
Lire la suite »
Meet the artist using AI to imagine better citiesWhat would the busiest street in your hometown look like with fewer cars and more people, and bike lanes, and light rail? Urban planning advocate Zach Katz cannot only tell you, he can show you — using the AI image-generation platform DALL-E.
Lire la suite »
Argo AI assembles panel of outside experts to oversee safety of its autonomous vehiclesArgo AI said the experts will help it deploy AVs safely.
Lire la suite »