Don't consider AI your friend, it'll hurt you: LinkedIn co-founder
What's the story
Reid Hoffman, the co-founder of LinkedIn and a prominent investor in artificial intelligence (AI), has warned against treating AI as a friend. In a recent episode of the Possible podcast, he said that no current AI tool is capable of being a true friend. "I don't think any AI tool today is capable of being a friend," he said. "And I think if it's pretending to be a friend, you're actually harming the person in so doing," Hoffman added.
Tech critique
Concerns amid AI companion trend
Hoffman's comments come as tech giants like Meta push for AI companions on their platforms. Meta CEO Mark Zuckerberg recently said he sees these chatbots as a solution to America's loneliness epidemic. However, Hoffman believes that such developments could be harmful if companies don't clarify what AI can and can't do. He also stressed the difference between companionship and friendship, saying the former doesn't necessarily involve mutual support or growth.
Friendship dynamics
Hoffman emphasizes the importance of mutual support in friendship
Hoffman further explained that true friendship involves two people supporting each other and growing together. He said, "It's not only, 'Are you there for me?', but I am here for you." This kind of connection is something no AI can provide, no matter how advanced it gets. He also praised Inflection AI's Pi assistant for being clear about its role as a companion and encouraging users to interact with real friends.
Industry standards
Calls for transparency and regulation in AI industry
Hoffman also called for more transparency and regulation in the AI industry. He said, "We as a market should demand it, we as an industry, all MPAs, should standardize around it." During a Senate testimony earlier this month, OpenAI CEO Sam Altman expressed similar concerns about AI developing personal relationships with children. When asked if he would want his own child to have an AI best friend, Altman said he wouldn't.