On Thursday, Meta’s AI chatbot mistakenly identified Joe Biden as the current U.S. president while also stating that Donald Trump was inaugurated in 2025. The gaffe sent shockwaves across the political and tech worlds, prompting Meta to issue swift corrections. This isn’t the company’s first misstep during the presidential transition. Reports of forced re-follows of Trump-related accounts and mysterious blocks on the hashtag #Democrat on Instagram have already put the social media giant under fire. Check out the conversation I re-created with Meta AI earlier today – it’s a fascinating example of machine learning; in particular, check out the way the answer changed the second time a question was asked – AI is always aiming to learn from previous interactions.
After this interaction, I went to go check Meta’s work. The Wikipedia page is, in fact, up to date – an example of how AI sometimes referencing out-of-date versions of pages. In this instance, Meta AI can only see the Wikipedia page from 2023, where Joe Biden is listed as president. Isn’t AI interesting?
Amid criticism, Meta’s recent internal changes, have only fueled scrutiny. However, the company isn’t backing down, with a $65 billion commitment to improving its AI systems. For many observers, the latest misstep underscores the urgent need for that investment—but whether Meta can navigate both technical and political challenges remains to be seen. What do you think of these worrisome AI blunders? Is $65 billion enough to make you trust AI? Let me know in the comments.
Want to get more content like this in your inbox? Make sure to sign up for my newsletter.