The CEO of Meta, Mark Zuckerberghas shared its intention to invest more in Artificial intelligence (AI), as well as using paid ads or content in interactions that use this technology in their services, such as the Meta AI assistant, to obtain benefits.
The US tech company launched its next generation of open-source large language model (LLM), Llama 3, last week, a model the company refers to as “best in class” with improvements in reasoning capabilities, code generation and more diversity in responses.
Alongside this, the company also began integrating its Meta AI assistant (based on Llama 3) into its social networks, thus adding new capabilities to access information from the web, among other features, and the Imagine function to create images from text in WhatsApp.
In this context, Zuckerberg has stressed that the new version of the Meta AI assistant and the new LLM Llama 3 bring the company closer to “building the world’s leading AI” and, although this technology can be accessed for free at the moment, the company has pointed out that there are “several ways” to create “a massive business” to obtain benefits by monetizing its assistant.
This was shared by Meta’s CEO in a call with investors, the highlights of which he shared on his Facebook profile, as part of the company’s first quarter earnings report, where he also highlighted that “it has been a good start to the year” for the company.
Specifically, Zuckerberg has announced that the initial deployment of Meta AI, which is currently available to users in some English-speaking countries and will be expanded to more languages and countries in the coming months, “is going well.” In fact, he has assured that “tens of millions of people” have already tried it.
Along these lines, he also expressed his satisfaction with the new models with 8B and 70B parameters of Llama 3, which he hopes will continue to improve “thanks to open source contributions.”
“Overall, I view the results our teams have achieved here as another key milestone in demonstrating that we have the talent, data and ability to scale infrastructure to build world-leading AI models and services,” he concluded.
Investing in AI and monetizing the results
With all this in mind, Zuckerberg has expressed his intention to “invest much more” in AI technology in the coming years, in order to build “even more advanced” models and AI services “on a larger scale in the world.” To do so, he has indicated that they will continue to focus on the efficiency of their services, improving the way models are trained and executed.
However, the executive also admitted that building these AI technologies will be a bigger challenge for the company than those it has undertaken so far and that, therefore, “it is likely to take several years.”
Still, Zuckerberg stressed that once the new AI services “reach the desired scale,” they will have “a solid track record of effective monetization.” With this, the CEO of Meta has indicated that there are several ways to create business with these AI services.
Specifically, he has referred to the possibility of expanding business messaging, introducing ads or adding paid content in interactions with AI. Similarly, he has proposed offering users a payment option to use larger AI models and access more computing.
AI in Meta’s social networks
In this regard, the CEO has also highlighted how AI is helping to improve participation on its platforms, which, he said, leads to showing more appropriate ads so that they “provide more value” and, therefore, users see more ads.
“So if technology and products evolve as we expect, over time each of them will bring enormous value to people and to our business,” he added.
In fact, Zuckerberg has also confirmed that around 30 percent of Facebook posts are now based on Meta’s AI recommendation system, a figure that has doubled in the past two years. Similarly, more than 50 percent of the content users see on Instagram is recommended by AI.
Finally, Meta has concluded that another of its long-term goals is the Metaverse, referring to its Ray-Ban Stories Augmented Reality glasses, which, as announced in March, have been enhanced with new multimodal AI capabilities, such as the ability to recognize places and landmarks to offer information about them.