Zuckerberg Critiques Current AIs, Meta Invests Heavily

⚡

Key Takeaways

1Mark Zuckerberg believes that current AI agents are too complex for the general public, despite Meta's massive investments.

2Meta reported a revenue of $56.3 billion, but the stock fell by 6% following an increase in investment spending.

3Meta is focusing on developing simple and accessible AI agents, unlike its competitors who are targeting more technical solutions.

💡Why it matters — Meta aims to make AI accessible to a broader audience, which could transform everyday interactions with technology.

Mark Zuckerberg Critiques the Complexity of Current AI Agents

During the first quarter 2026 earnings conference, Mark Zuckerberg expressed skepticism about the user-friendliness of current AI agents. According to him, these technologies are still too complex for widespread adoption by the general public. Nevertheless, Meta has decided to increase its investments in research and development, reaching up to $145 billion this year, with the aim of creating more accessible and efficient AI agents.

On April 29, Meta released its financial results, announcing a revenue of $56.3 billion for the quarter, a 33% increase compared to the previous year, with a net profit of $26.8 billion. However, despite these impressive financial results, Meta's stock fell by more than 6% after the announcement of increased capital expenditures, rising from $115-135 billion to $125-145 billion for the year. Investors remain skeptical about how these investments will translate into tangible revenue.

Zuckerberg questioned how to develop a more refined and simplified user experience for these AI agents. He emphasized that the quality of the product is more important than meeting a specific launch date.

Meta Faces Competition from Cloud Giants

Unlike Amazon, Microsoft, and Google, which generate direct revenue by selling their AI cloud services to businesses, Meta is focusing on developing AI agents aimed at the general public. However, monetizing these agents remains a challenge. Susan Li, Meta's Chief Financial Officer, noted a significant increase in interactions with professional agents, rising from one million weekly conversations in January to over ten million by the end of March. These tools are generally free for most businesses, raising questions about Meta's business model.

Mark Zuckerberg also pointed out that while many AI agents exist, few are simple enough to be recommended to his own mother. He insists on the need to create user experiences that are both optimized and simplified.

OpenClaw: An Example of Technical Complexity

The AI agent OpenClaw, developed by Sam Altman, exemplifies the simplicity challenges that Zuckerberg aims to overcome. OpenClaw is open-source and self-hosted, requiring local installation, terminal access, and manual system configuration. It integrates platforms like WhatsApp and Telegram and functions as an adaptive AI with memory. This technical complexity limits its adoption to a few million users, while Meta claims 3.56 billion daily active users. Meta's goal is to develop an AI agent that requires no prior setup, documentation, or complex steps.

Analysts at Loop Capital recently highlighted that Meta's massive investments could be seen as an attempt to catch up in the AI space. In response, Zuckerberg introduced Muse Spark, the first model from the Meta Superintelligence Lab, to dispel these doubts. This model is already integrated into discussions on Facebook, Instagram, and WhatsApp, as well as on the Meta AI app and website.

Muse Spark and Meta's Ambitions

Since the rollout of Muse Spark, Meta has seen a double-digit increase in Meta AI sessions per user. Over eight million advertisers are using at least one generative advertising tool from Meta, with conversion rates rising by more than 3% for those utilizing video generation. Zuckerberg also mentioned that users might be willing to pay for premium or high-capacity versions of these agents.

Mark Zuckerberg stated during the conference that Meta's goal is not just to deliver Meta AI as an assistant, but to provide agents that understand users' goals and work tirelessly to help them achieve those.

Meta Distances Itself from the Coding Agent Race

While companies like OpenAI, Anthropic, and xAI invest in coding agents, Meta is choosing a different path. OpenAI launched Codex, Anthropic is focusing on Claude Code, and xAI acquired Cursor for $60 billion. However, Zuckerberg insists that programming is just one aspect among many of personal development. Muse Spark, for example, supports multiple modalities, tool usage, visual reasoning, and multi-agent orchestration.

Meta also acquired Manus, a Singapore-based startup specializing in autonomous agents, for over $2 billion at the end of 2025. Manus focuses on the autonomous management of complex tasks such as research, analysis, and writing, charging businesses on a subscription basis, generating an annualized revenue of over $125 million.

No specific date has been announced for the launch of Meta's personal and professional agents, but the company continues to work on solutions that could transform how users interact with technology on a daily basis.