The use of generative AI platforms continues to grow at high speed

Meta CEO Mark Zuckerberg recently indicated that Meta AI, his assistant capable of answering questions asked in everyday language or generating images, now had 400 million monthly users.

Launched in April, Meta AI has been integrated into the social networks Facebook and Instagram, as well as WhatsApp and Messenger, but it is also accessible via a dedicated site.

It is based on the Llama 3 language model, developed by Meta and built through the accumulation of massive amounts of data to allow it to choose the best answer to a query.

Its competitor OpenAI had previously indicated to several American media that its generative AI interface ChatGPT was used by more than 200 million Internet users each week, double the number of weekly users announced in November.

Launched in November 2022, ChatGPT was the first “chatbot” to revolutionize the use of generative AI by the greatest number of people.

At the end of July, Microsoft reported that 77,000 business customers were using the group’s AI assistant, called Copilot. Asked to obtain an updated number of users, Microsoft did not immediately respond.

If ChatGPT and Copilot exist in a paid version, the three major generative AI interfaces all offer a free plan.

Meta has, for the moment, chosen not to offer its interface in the United Kingdom, the European Union and Brazil.

In mid-July, the group announced the postponement of the launch of the new version of its Llama 3 model in the EU as a precaution, claiming to lack visibility regarding the interpretation by the European authorities of the general data protection regulation ( GDPR).

The development of Large Language Models (LLM) requires the use of an immense amount of data, some of which relates to EU users, which could contravene the provisions of the GDPR.

Among the most used generative AI interfaces is also Gemini, from Google.

Source : AFP