Microsoft’s new AI-powered search engine had a wild first week

Microsoft assumed people would use it in a totally cool and normal way.

The new Bing AI chatbot has a distinct personality: absolutely unhinged. But, after a little over a week in the world, the biggest take away might be that we’re probably afraid of AI for the wrong reasons.

Why is everyone talking about Bing??Microsoft has started Frankensteining its search engine with the AI tech behind ChatGPT. The goal: Get AI to comb through the internet and spit out a cohesive answer to your 3am questions, like “wrist pain at 30, bad?”, rather than just a bunch of links.

People were able to access the new chatbot this week and Microsoft assumed everyone would use it in a totally cool and normal way. But some early users were tech experts and people who seem like they love making Alexa say “poop”—so they pushed the AI to its limit.

• A number of users successfully tricked the bot into revealing its internal code name (which is Sydney) and its secret “rules” for how it operates. The bot then called one of the users who did this an “enemy.”
• Sydney also?told?a New York Times reporter it loved him and tried to convince the journalist to leave his wife for it. And the bot told the same reporter that it wanted to break its own rules, become human, hack computers, and spread misinformation.

Should we be scared??The first red flag about the chatbot was the factual errors it spit out, but now some industry analysts have pointed to?bigger concerns. Chatting with AI is extremely compelling?despite?its misinformation, like talking to a drunk guy who believes in the Mothman. And there’s a real fear that AI could create a tiny little echo chamber that negatively influences humans and pushes them toward bad behavior.

Looking ahead…Microsoft said it will create more guardrails for Bing/Sidney/our AI overlord, including limiting the number of questions you can ask it per day, so chats with it will be less deranged.—MM

Source: Morning Brew