Unlike other chatbots and visual generators, Meta’s AI assistant is a free tool built into the apps billions of people use every day.
In a response to OpenAI’s ChatGPT, Mark Zuckerberg’s Meta has brought AI chatbots to the company’s most popular applications such as Instagram, Messenger and WhatsApp, but users need to be careful as the algorithm still has many errors. .
Unlike other chatbots and visual generators, Meta’s AI assistant is a free tool integrated into apps that billions of people use every day, making this the most proactive effort from a technology company to bring AI to the mainstream.
“We believe Meta AI is currently the smartest AI assistant you can comfortably use,” Mark Zuckerberg, the company’s chief executive, wrote on Instagram.
However, the New York Times said that Mark Zuckerberg’s chatbot still makes many errors, especially for search tasks.
The New York Times warns users to be careful when using Meta’s AI chatbot. Photo: Reuters
Meta representative said that similar to other AI systems, because the technology is still new, it cannot always return accurate responses. Currently, AI chatbots are integrated by default within Meta applications and users do not have the option to turn off the service.
The parent company that owns Facebook says its AI chatbot can be a replacement for web browser search engines. However, even with the simplest queries such as searching for recipes, airline ticket prices or fun weekend activities, Meta’s algorithm cannot provide accurate responses.
However, this AI assistant is not a complete failure, one of the bright spots is the ability to edit text. For example, when a user asks the chatbot to rewrite a long paragraph, it can omit unnecessary words or flexibly change between active and passive sentences.
Another bright spot is the speed at which images are created. Meta’s AI is much faster than other image generators like Midjourney – the service can take more than a minute to return results.
Currently Mark Zuckerberg’s company is accelerating in the field of AI to catch up with competitors like OpenAI. The social networking giant has planned to integrate the latest version of the most advanced Llama 3 large language model into its AI chatbot in the next few months.
This year, the smart glasses models that the company cooperates with Ray-Ban will also receive an AI update, allowing users to identify objects and answer related questions.
Previously, Meta announced a partnership with Google to include real-time search results in chatbot responses, complementing an existing agreement with Microsoft’s Bing engine.
Meanwhile, AI assistants have also been expanded beyond the US market, including dozens of other countries and territories such as Australia, Canada, Singapore, Nigeria and Pakistan. However, this technology is not yet “made in Europe” – which has strict privacy regulations and requires disclosure of data used in algorithm training.
Mark Zuckerberg said that the largest Llama-3 model that the company is developing has 400 billion parameters (smaller versions include 8 billion and 70 billion parameters). To alleviate contextual issues, Meta says it uses “high-quality data” so the model can distinguish between different nuances.
The company revealed seven times more data input to train Llama-3 than it used for Llama-2.
(Theo NY Times, Reuters)
