No one likes the Ai Glitchy app from Zuckerberg

Obviously, throwing billions of dollars on a problem cannot solve everything.

Almost six months after its beginnings, the AI -oriented Meta -oriented application is still fighting against inconsistencies and persistent gaps, casting a doubt about the high ambitions of the company for artificial intelligence.

It was an approximate apprenticeship curve for the efforts of Mark Zuckerberg to join the AI race after having lagged behind an already frequented field and spending tens of billions of dollars to catch up. Until now, the most meta must show because it is a glitchy chatbot, an increasing choir of irritated users and frustrated shareholders who would like to have something to show for this type of capital expenditure.

Now, market observers and consumers wonder what Meta can deliver exactly if its flagship offer of AI remains unequal.

Launched in April 2025, the autonomous application was a late entry into IA space, arriving about two and a half years after the introduction of Chatgpt, the largest competitor of Meta in space and by far the most commercially recognizable brand for AI today.

Meta AI also adopted a different approach: her bot ai tries to combine the cat, the creation of images and a public flow with content generated by the user.

Success or Miss was mainly missed

This tack seems to fail. Internet users have expressed their frustration with its unpredictability and limited relevance, saying that it looks less like a polished product and more like an early prototype that fights to live up to the new AI -focused vision.

Since its launch, criticisms and users have expressed concerns about social media, highlighting buckets, strange interactions and lack of personalization.

The application flow flow, intended to inspire conversations and present the creative uses of AI, often displays images generated by the user, reducing its appeal as a dynamic social center.

Meanwhile, its cat function, which supposes allegedly on user preferences, frequently constitutes false information – a problem called “hallucination” in the AI jargon – which questions its reliability.

The reception was at best lukewarm.

A Reddit article entitled “Who hates Meta Ai?” has collected thousands of votes, commentators rejecting him as “here anyway” and “nobody has ever asked”.

Depending on the company’s own words, the application is still in its infancy. “This is only the first of the many stages,” said a spokesperson for Meta in Bloomberg, promising current updates while the company invests massively in AI talents and infrastructure.

The success of the program and its importance for the future of Meta cannot be overestimated.

Zuckerberg has made AI an cornerstone of Meta’s future, promising hundreds of billions of dollars to development and actively recruits the best rivals researchers like Apple and Openai. Meta even Open A Open, its large language model, Llama, positioning itself as a leader in AI research, but its consumption tools, including the Meta AI application, are still far from polished.

The aspirations of the meta go beyond simple chatbot interactions

Zuckerberg said that he considered AI as a “personal superintendent” which allows individual users, not only as an entertainment tool. The company said that it possibly wanted to integrate AI into its application portfolio, including Facebook, Instagram and Messenger and the equipment.

But the current experience does not become again from this vision, because the way in which the meta is deciding and competing in the data it receives is to thwart internal efforts to advance its AI program. Conversations with AI are compartmentalized on applications, and it does not keep memory or context from previous cats, unless they explicitly programmed to do so, limiting its usefulness and personalization.

More worrying, sometimes hallucinating AI, making details that users can believe to be true. It can also become quite bizarre, with many images generated by AI and text extracts that he uses to flood the flows with content generated by the user ranging from inappropriate to bizarre and bizarre.

This was aggravated by the Meta grapple with automated equilibrium and moderation.

META’s objective to exploit the potential of AI applications for productivity and entertainment remains largely ambitious. For the moment, the most visible consumer product of the company remains a work in progress, far from carrying out the high promises of Zuckerberg of a future powered by AI designed for “individual empowerment”.

While Meta continues to develop its AI technology, it remains to be seen if the company can transform this emerging application into a personal assistant really useful and trustworthy, or if it will continue to fold its tests in real time and to become a edifying history of moving too quickly.