“It’s understanding that anything you get out of one of these large language models, Meta’s included, could very well be fake, incorrect,” Dave Hatter, a cybersecurity consultant with Intrust IT, said. “It could have bias, you know, all kinds of things like that baked into it.
“The question is, will it produce results as good or better than these other platforms? I’ve seen conflicting information on that, too,” Hatter said. “Some are saying they’re better than some of the competitors out there. Others are saying not so much.”
Hatter says Meta AI is based on Meta’s open-source AI model, called LLAMA.
Here’s how it works: In your Facebook or Instagram app, the search bar now reads “Ask Meta AI Anything.” You can ask for things such as recipes, help with writing a letter or to create a picture.
Plus, it does regular in-app searches, too.
You might also notice the “Ask Meta” button on certain Facebook posts. If you click it, a window will pop up, where it says, “Think of me like an assistant.”
For those who may want to get rid of the newest feature, there’s no clear way to do so. The best way to avoid using Meta AI is to just use the search tool for regular searches and, if you do accidentally hit AI, you can exit out of it.
