Meta Platforms Inc. is taking significant steps to prevent inappropriate replies from its new artificial intelligence (AI) products, including chatbots. The company's president of global affairs, Nick Clegg, has assured that they have invested in comprehensive testing to make their AI tools safer.
New AI products by Meta
Meta Platforms Inc. has recently released a variety of new AI products, including a chatbot designed to converse like a detective while appearing as Paris Hilton, and another character, the 'Dungeon Master,' modeled after Snoop Dogg. These amusing additions are part of a serious effort by Meta to integrate AI tools, capable of responding to user messages, into apps used by billions monthly.
Meta's commitment to responsible AI use
Nick Clegg, president of global affairs at Meta, is committed to ensuring the responsible use of AI by the company. This commitment comes in the wake of years of criticisms against Meta for its alleged lack of sufficient safeguards on its products. Clegg maintains that they have invested thousands of hours in 'red teaming' or stress testing their AI tools to make them as safe as possible.
Challenges with AI's unpredictability
Meta is facing new challenges with its AI tools, which could potentially deviate from their scripts or be a source of misinformation. The company is putting in significant effort in addressing these issues, maintaining that regular retraining of the AI systems is part of their strategy to mitigate such risks.
Nick Clegg was brought onboard by Meta in 2018 to oversee global lobbying activities. His appointment came when the company was tackling crises related to privacy concerns and allegations of election interference. His task now involves not just lobbying but also addressing the challenges and responsibilities that come with developing AI systems.
Clegg has highlighted the important role of transparent and open-source licenses in maintaining ethical use of AI. According to him, the use of such licenses for certain datasets and large-language models, which form the technological base for AI chatbots, is one way Meta aims to ensure the ethical usage of its AI technologies.