Meta Platforms is gearing up to introduce AI-powered chatbots with distinct personalities in a move aimed at improving user engagement and retention on its platforms.
The chatbots are reportedly scheduled for a September release and will be responsible for offering recommendations to users, providing search functions and just being “a fun product for users to play with.”
The Lincoln chatbot
The chatbots, which staffers at the social media giant have dubbed “personas” are expected to hold humanlike conversations.
As reported by the Financial Times, sources close to the development indicated the new chatbots will exhibit different personalities, and mimic high profile individuals. One of the chatbots will speak like former President of the United States Abraham Lincoln, according to reports. The other chatbot will help users with travel information in the style of a surfer.
The report further highlights that Meta may also explore an avatar chatbot in the metaverse, according to a person close to the developments.
“Zuckerberg is spending all his energy and time on ideating about this,” said the source.
During the company’s earnings call during the last week of July, Meta CEO Mark Zuckerberg said the company envisaged “agents that act as assistance, coaches or that can help you interact with businesses and creators and more.” He added the company will highlight more about its product roadmap for AI next month at its Connect developer event.
Apart from boosting engagements on social platforms, the Financial Times reports that the chatbots may help the company collect data on users’ interests, which will help provide relevant content and adverts to users, boosting its earnings which comes mostly from adverts. During the last week of July, the company reported strong second-quarter advertising revenue, which was topping Wall Street projections.
However, there are also concerns about user privacy as this may also expose users to potential “manipulation,” according to Ravit Dotan, an AI ethics adviser and co-founder of the collaborative AI Responsibility lab at the University of Pittsburgh.
“Once users interact with a chatbot, it really exposes much more of their data to the company, so that the company can do anything they want with that data,” she said.
Data privacy is one of the concerns that have been raised by stakeholders concerning the use of AI and has also been topical in discussions about regulating the industry.
The company is therefore expected to go under scrutiny from experts looking at possible signs of bias and misinformation. In 2021, Meta released BlenderBot 2, but was found to be spreading misinformation and dangerous information.
However, close company sources have revealed the company may build technology that screens questions from users and ensures the responses are appropriate.
Race to stay ahead
Like Meta, other tech companies have also introduced chatbots that have personalities and can have human-like conversations. Character.ai is one such firm that uses a large language model (LLM) to generate chat in Tesla boss and Twitter – now X owner Elon Musk’s style. The bot also converses like the Nintendo character Mario.
Now, the plan by Meta comes at a time when competition is already coming from other social media platforms like TikTok, pushing the company to innovate to retain its current market share while attracting other users to its different platforms.
Its latest platform and Twitter rival Threads has lost more than 50% of its users after achieving a record of signups days after its launch in July.
Tech companies are also in a race to leverage AI in their businesses, which was spurred by the Microsoft-backed OpenAI’s viral chatbot ChatGPT, which launched in November 2022.
Last month, Meta released an AI model known as Lama 2 for commercial use as the race for AI supremacy continues. According to Bloomberg, Apple is also working on its AI chatbot to rival OpenAI’s ChatGPT and Google’s Bard.