Meta is deploying itsAI Studiotool to more USInstagramusers this week, giving them a chance to build their own AIchatbots. While the bots can be used for your own entertainment, they can also act as doppelgangers, accessible by your followers and/or the general public – depending on the privacy settings you choose. In fact Meta is promoting the idea of chatbots as a tool for creators, providing a quick way of answering common questions and linking to past content, with customization options such as who they’ll reply to and the topics they should avoid.

How to use Gemini in Google Messages

It’s much easier to access Gemini in Google Messages than a separate app, but what it can actually do might not be useful for everyone.

You don’t have to be a creator though – anyone in the US can customize a bot with its own name, personality, tone, avatar, and tagline. There’s an assortment of prompt templates if you want to save time, although you’ll need to put in some effort if the bot serves a niche purpose like teaching people how to cook.

Typing to Gemini in Google Messages on a Pixel 7a.

All bots are based on Meta’sLlama 3.1AI model, which was launched as an open-source project just last week. That model is already a little controversial, since it’s only free for commercial purposes up to 700 million users. The policy may be meant to deter rivals from beating the company at its own game.

How will I know if I’m talking to a bot?

Bots are clearly identified with an AI label. That said, Meta is hoping doppelgangers can come close to the real thing – creators can have them use previous comments and captions as source material, or even the transcripts of Reels. In the past the company has experimented with bots mimicking celebrities like Snoop Dogg, though not to great success.

Some creators could be leery about adopting AI, since it’s impersonal and poses the risk of offending fans. Indeed AI Studio’s disclaimers warn that messages produced by AI “may be inaccurate or inappropriate,” referring to the fact thatgenerative AIis prone to hallucinating – that is, making things up when it interprets data poorly. Those mistakes could damage the reputations of the many celebrities and influencers on Instragram.

An Instagram chatbot created in AI Studio.