Chatbots: Why does White House want hackers to trick AI? - BBC News
Technology reporter
Under the spotlight are large language models, those helpful chatbots like OpenAI's ChatGPT and Google's Bard. Meta, Google, OpenAI, Anthropic, Cohere, Microsoft, Nvidia and Stability have been persuaded to open up their models to be hacked to identify problems. The prize is a powerful piece of computing kit, a graphics processing unit, but perhaps more important according to Dr Chowdhury will be the "bragging rights".One of the challenges asks hackers to get a model to hallucinate or invent a fact about a political person or major figure.
Danmark Seneste Nyt, Danmark Overskrifter
Similar News:Du kan også læse nyheder, der ligner denne, som vi har indsamlet fra andre nyhedskilder.
AI chatbots like ChatGPT could be security nightmares - and experts are trying to contain the chaosNew research reveals crucial vulnerabilities in AI chatbots
Læs mere »
House prices are falling, so why can't I afford a home?Figures this week showed house prices falling, but mortgage costs remain high.
Læs mere »
Meet the magician making waves in an industry dominated by white menMichael Areola, or King of Aces, is a magician from Accrington that's breaking down barriers in the entertainment industry
Læs mere »
The ‘Beautiful Game’: Why we can't underestimate the role of beauty in women’s footballFYI, women can look good and score absolute belters ⚽️
Læs mere »
What the hell are the police thinking? Why wasn't Rishi Sunak's house guarded, says Nick FerrariYou can't blame great Greenpeace, although I don't hold with their tactics. You can't blame them. They've got their point across.
Læs mere »
Derry: Vandalism of rainbow crossing treated as hate crimeWhite paint was splashed across the rainbow pedestrian crossing at the Foyle embankment in Derry.
Læs mere »