AI in My WhatsApp? At First, I Said “No Thanks”.
Initially, I thought, “No, thank you.” Who needs more AI options when we already have enough to throw and give away? Until a friend told me they used it for quick inquiries… and I liked it!
AI on Facebook and Instagram: Not for Me
As for AI on Facebook and Instagram, I’m definitely not interested. They’ve made so many changes, and the advertisement in networks has become so complicated—when they promised it would be revolutionary—that for now, their advancements don’t appeal to me. But there is something that does interest me: non-generative AI, not the kind we mere mortals use to create memes or school tasks, but the kind working on ambitious projects behind cameras.
Meta’s AI: Not Just Anything
Meta’s AI (the company of Mark Zuckerberg encompassing Facebook, Instagram, WhatsApp, and other projects) is no ordinary thing and didn’t arrive late. It has been in development for over a decade, backed by substantial investments and plenty of controversy.
Open Source AI
This is common knowledge among the informed. In 2013, Facebook established FAIR (Facebook AI Research) under the direction of Yann LeCun, a pioneer in deep learning. Their goal wasn’t to help us make photos of The Simpsons but to create projects in language processing and deep learning models.
Thus, Facebook began experimenting with bots in Messenger and later with automatic translation systems. If your Facebook suggests friends, that’s AI too. But in 2022, they went further: Meta surprised everyone with Cicero, a system capable of negotiating and persuading in the game Diplomacy. In 2023, they launched LLaMA (Large Language Model Meta AI), an open-source language model for researchers and developers.
Here’s the difference: while OpenAI (ChatGPT) or Google (Gemini) keep their models closed, Meta bets on openness. For Zuckerberg, AI should be democratic and accessible.
AI with Social Impact
What does Meta’s AI do? In 2024, for example, they organized a hackathon in London with over 200 developers proposing solutions for people with reduced mobility, pharmacists, and public health projects. Their models have also been used to map communities without electricity or potable water and are integrated into economic and social initiatives.
In his 2025 AI manifesto, Zuckerberg wrote: “Superintelligence has the potential to usher in a new era of personal empowerment where people have greater ability to improve the world in the directions they choose.”
Musk vs. Zuckerberg
Meta’s AI is also understood through the public rivalry between Mark Zuckerberg and Elon Musk, which began in 2014 and continues to this day.
While Musk insists AI is dangerous, Zuckerberg defends its potential to change the world. In 2017, Zuckerberg said, “I am very optimistic about AI. Those who try to generate catastrophic scenarios… quite simply, I don’t understand them. It’s negative, and in a certain sense, quite irresponsible.”
Musk, on the other hand, has issued dramatic warnings: “With AI, we summon demons… In all those stories with the guy and the pentagram, he believes he can control it. It never works.”
Who’s right? That’s another story… We’ll detail soon how Elon Musk’s AI works.
The “Super-Intelligent” Team
Here’s the most controversial part. In February 2025, Meta fired 3,600 employees. It wasn’t the first time: between 2022 and 2023, they cut 20,000 positions globally.
Zuckerberg justified these dismissals as “performance terminations” (for low performance). However, analysts pointed out that the real reason was to free up resources for hiring more AI experts.
In an interview, Zuckerberg stated, “We will reach a point where all the code of our applications will be written by AI engineers instead of human engineers.” He then assured that AI would replace many mid-level software engineers, and most of Meta’s code would be generated by automated systems.
This strategy is serious. This year, Meta hired high-level personnel from OpenAI, and according to rumors, they offered bonuses of up to $100 million to attract top researchers and engineers.
Parallel to this, they invest millions in hardware, such as their own GPUs, servers, and data centers designed for AI.
Because Meta isn’t just betting on research. All their AI is tested on their platforms (Facebook, Instagram, WhatsApp). How can you detect it? It’s what recommends friends, moderates your comments, suggests content generation, manages your ads, and handles customer assistance chats.
Have you noticed? Yes, Meta’s AI has been among us for a while, though we may not always realize it.
How will all this change the current AI landscape? Because if there’s one thing we can be sure of, it’s that Meta not only wants to compete but also redefine the game.
Testing AI in WhatsApp?
If you want to experiment with Meta’s AI on your phone, the next time you open WhatsApp, try its search bar:
- Ask for summaries of voice notes or long conversations.
- Request message reminders (“When did I have dinner with my mom?”).
- Organize the content you share (recipes, memes, photos).
- And my favorite: make quick inquiries like currency exchange rates, weather, or song lyrics without opening Google. Just don’t expect advanced detail levels. I couldn’t get the Meta AI itself to explain how it works!
Key Questions and Answers
- What is Meta’s AI? Meta’s AI, developed by Mark Zuckerberg’s company encompassing Facebook, Instagram, WhatsApp, and other projects, is not just any AI. It’s been in development for over a decade, backed by substantial investments and controversy.
- How is Meta’s AI different from others? Unlike OpenAI (ChatGPT) or Google (Gemini), Meta bets on open-source AI, making it democratic and accessible.
- What impact does Meta’s AI have? Meta’s AI has been used for various social initiatives, including solutions for people with reduced mobility, pharmaceutical projects, and public health endeavors.
- Why is there a rivalry between Zuckerberg and Musk regarding AI? The public rivalry between Mark Zuckerberg and Elon Musk stems from differing views on AI’s potential—Zuckerberg sees it as transformative, while Musk warns of its dangers.
- How is Meta integrating AI into its platforms? Meta tests all its AI on its platforms (Facebook, Instagram, WhatsApp), influencing user experiences through recommendations, content moderation, ad management, and customer assistance chatbots.
