Meta wants a seat at the table for AI chatbots so is launching a string of new, AI-powered bots across its messaging apps, including WhatsApp, Messenger, and Instagram. They're live in beta with select U.S. users today.
There's Meta AI, which the company is talking up as a virtual assistant: from its own hands-on piece on The Verge, it seems highly similar to OpenAI's ChatGPT and Anthropic's Claude 2. Meta AI — which will shortly be making its debut on Meta's new Quest 3 VR headset, Meta CEO Mark Zuckerberg said onstage at Meta's Connect conference today — can help plan a trip with friends in a group chat, answer general-knowledge questions and search the internet across Microsoft's Bing to provide real-time web results.
M reminds this reporter of Facebook's virtual assistant, which was destined to fail as it could, among other things, make plans, automate payments, and even place a phone call on a user's behalf. This in the case of M, which existed only for about 2,000 people living in California and suffered from development neglect.
But Meta AI, powered by a "custom-made" large language model, is set to get a much broader release. And whereas M relied in part on a team of humans to fulfill requests, Meta AI is totally automated, with a model that can refer back to previous conversations and that's been tuned to give "very concise" answers. Maybe that will be the difference.
Meta AI can be called up in any conversation, Meta says. And just like ChatGPT, it can also generate watermarked images using an offshoot text-to-image model called Emu, developed by Meta's research division into artificial intelligence.
Beyond Meta AI, Meta today announced a slate of "AI characters"- essentially chatbots fine-tuned to capture particular personalities and emulate celebrities, from Kendall Jenner and Dwyane Wade to MrBeast, Paris Hilton, Charli D'Amelio and Snoop Dogg. More bots, like Bear Grylls, Chloe Kim and Josh Richards, are on the way. Like Meta AI, the chatbots-living in Meta's messaging apps, with illustrated or real-life profile images-will animate avatars subtly based on the conversation as you chat with them.
This all sounds interesting in theory, but some very recent history has shown that even the most seemingly intelligent AI chatbots go off the rails and simply manufacture text and miss important points in conversations.
Meta says that it spent 6,000 hours finding problematic use cases and recruiting employees to interact with the models internally before releasing them. And it claims that it's developed new tech to catch and take action on content that violates its policies; prevented chatbots other than Meta AI from searching the web; and made its new chatbot features available to researchers through its bug bounty program.
We'll have to see, however, if they stand up to the test of real-world life.