Mark Zuckerberg Invests Billions of Dollars to Achieve the Ultimate AI Goal

Superintelligence and Artificial General Intelligence: Ambitious Concepts

Leading AI companies are currently moving towards an ambitious goal known as "superintelligence", a hypothetical AI system that significantly surpasses the cognitive capabilities of the brightest human minds in almost all intellectual fields, including scientific creativity, general wisdom, and social skills. At the same time, major companies in this field are striving to develop "Artificial General Intelligence" (AGI), a theoretical form of AI that aims to achieve a level of intelligence comparable to humans, enabling it to understand, learn, and apply information across a wide range of intellectual tasks, thereby performing any mental task a human can.

The term "superintelligence" might just be a marketing strategy to attract more investments, or it could represent the logical next evolution for AI models that some already consider to be at human-level intelligence.

Meta's Strategy Behind Superintelligence and the Talent Race

Specifically, "Meta" is focusing its AI division's efforts around the goal of superintelligence through its new "Meta's Superintelligence Labs". In recent months, Mark Zuckerberg has intensified his efforts to attract talent in this competitive sector, with reports indicating he offered compensation packages up to a billion dollars to recruit the best AI engineers in Silicon Valley.

This spring, Mark Zuckerberg realized that "Meta" was falling behind in the overall AI race. Meta had just launched the latest version of its large language model, "Llama", which aimed to compete with systems like "ChatGPT" from "OpenAI" and "Claude" from "Anthropic". However, upon its release, Llama did not achieve the expected success, and it became clear that Meta was not at the forefront of AI companies.

Meta's Transformation: From Metaverse to AI Leadership

While many associate "Meta" with the parent company of Facebook, Instagram, or WhatsApp, Meta increasingly sees itself as a leader in the field of AI. Given its lack of leadership in this area, Zuckerberg immediately took action. He began personal recruitment and built a new secret team now known as "Meta's Superintelligence Labs".

When talking about "Meta", the "metaverse" is often seen as a weakness, considered a money-draining project that isn't a significant source of revenue, and it hasn't fulfilled the dream Mark Zuckerberg once envisioned. However, Meta's AI ambitions are somewhat different. In the days when the company was still known as "Facebook", Meta had a presence in the field of AI. It brought in Yann LeCun, a prominent and award-winning thinker in this field.

The Impact of ChatGPT and the Acceleration of the AI Race

But when "OpenAI" launched "ChatGPT", it caused widespread astonishment, even among AI researchers at Meta. They were working on building their own large language models, but their approach was still purely academic. It wasn't presented as a consumer product in the successful way OpenAI did with ChatGPT. Thus, the "AI race" began, and it accelerated significantly in April with what many inside Meta considered a failure in the latest version of Llama.

The Talent War and the Shift Towards Deductive Reasoning

We are currently witnessing a fierce talent war, and "Meta" is creating a new organizational structure for its AI teams. This effort requires an investment of billions of dollars, and Meta has the financial capacity to spend it. The company's leadership perspective is that they are betting heavily on closing the talent gap. They believe there's only a small pool of talent capable of maximizing this expenditure to develop competitive models. Many are focusing on deductive reasoning, meaning they are specialists who can help build models that think step-by-step, rather than the probabilistic approach that predicts the next word in a sentence.

Many current chatbots seem to understand what you're saying when you type a question, but they are actually just predicting. They have ingested vast amounts of data and are capable of predicting the next word in a sentence. What "Meta" and "OpenAI" are trying to do is give them the ability for true reasoning, so they can deduce on their own.

The Impact of Meta's Strategy on the Competitive AI Landscape

It's too early to say whether this approach will work. What has already been successful are Mark Zuckerberg's efforts to attract some of the biggest names in the field. They are recruiting top talent from competitors. However, we also know that there are people who have rejected Mark Zuckerberg's compensation offers that rival sports teams. Nevertheless, this rejection does not come from Wall Street; Meta's stock has risen by about 30 percent year-to-date. Clearly, there's significant enthusiasm from the investor community.

It appears "Meta" is restructuring itself or creating entirely new structures, to play the long game in superintelligence. Everyone, indeed everyone, has been forced to respond to the aggressive recruitment campaign launched by "Mark Zuckerberg". His efforts have set a new standard for compensation across the industry. We've seen other companies like "OpenAI" forced to offer their employees significantly higher packages.

We see this even in the language and public statements of CEOs. "OpenAI" has stated it will spend trillions of dollars on the AI race. Other companies are loudly talking about the amounts they will spend. But we're talking about the biggest players here – not startups that just secured some venture capital and have no revenue. "Meta", "Google", and even "OpenAI" are in a different position because they have massive cash reserves to spend.