Meta Aims for Open-Source AGI

Greetings, intelligists!

As always, there is much to talk about, but today, we are focusing on one of the most compelling and potentially significant headlines of the past couple of days.

On Thursday, Mark Zuckerberg, CEO of Meta, announced that he has set his company’s sights on the ambitious goal of developing artificial general intelligence (AGI).

AGI is a hypothetical AI milestone in which the technology achieves a state of comprehensive, complete, and possibly autonomous capabilities. AGI might effectively imply the existence of AI that can far surpass human intelligence.

Furthermore, Zuckerberg announced via Instagram that the company is transitioning to make AI a top priority starting now. Another of the most startling revelations is that Meta is investing at least $9 billion in 2024 alone, just on Nvidia H100 graphics processors, or GPUs. This is part of a transition plan in which Meta will gradually go all-in when it comes to AI.

At the same time, the Meta boss made it clear that the company is pushing Llama forward. Llama 2, the open-source LLM, made big headlines late last year as a viable alternative to ChatGPT. The Facebook co-founder also stated that Llama 3 was in the works, which could produce another field day for open-source AI developers who can modify and iterate on the AI model.

For many in the AI world, Meta has been commended for going the open-source route. The decision to open-source their model starkly contrasts with OpenAI and Google, who have decided to keep their source code private from the public and developers. Open-source software advocates emphasize that transparent and collaborative development is the only way to create a safe and ethical way forward, especially in AI.

While Zuckerberg was emphatic in his support for Meta’s current open-source projects, he was less clear when implying that Meta's AGI advancements could be open-sourced.

This bold strategy aligns with Meta's restructuring, integrating its AI research group FAIR with its generative AI product teams. The company’s objective is now to push Meta's AI breakthroughs directly into the hands of its billions of users.

However, Zuckerberg's pursuit of AGI raises crucial questions about control and ethical governance in AI development. Despite Meta’s significant level of commitment to open-source thus far, it is unclear whether they will stick to this course in their AGI pursuit.

In addition, data risks and privacy are significant concerns. Meta’s past privacy record is not exactly exemplary. When you stop to consider that AI requires vast amounts of data, it is impossible to ignore that the source of that data could very well be Meta users. Instagram, WhatsApp, and Facebook have over 3.5 billion users worldwide. Putting AI into the Meta ecosystem could imply billions of users providing Meta with extremely valuable data.

Ultimately, it is necessary to question whether Meta is truly interested in democratizing access to advanced AI technologies. The company is a for-profit enterprise, and there will always be financial motives behind its actions.

Amidst these bold initiatives, Zuckerberg reaffirmed Meta's commitment to the Metaverse. For many in tech, the Metaverse became a joke about the extremes and excesses of technological innovation. Despite this, it seems that Zuckerberg views it as a platform where AI can be developed and interact with humans in a virtual world.

Overall, these announcements from Meta are exciting. Meta’s products, whether you like them or not, are an integral part of the technological lives of billions of people worldwide. Many people rely on platforms such as WhatsApp, Facebook, and Instagram to keep in touch with their family and friends, grow their businesses, network, and more.

Now, the tech giant will throw their significant weight behind AI development. For some, this news heralds a changing future in which AI impacts humanity in ways we never imagined. For others, it is troubling to see a mega-corporation take control of some of the most powerful and rapidly developing technology known to humankind.

One thing is beyond doubt: we are witnessing a keystone moment in human history, and there might be no turning back from the new directions of artificial intelligence.