In a bid to cement its presence in the ever-evolving world of artificial intelligence, Meta, the tech behemoth formerly known as Facebook, has set its sights on a monumental task: training a new AI model that can stand shoulder to shoulder with OpenAI’s latest GPT-4 chatbot. According to a recent report from The Wall Street Journal, Meta is leaving no stone unturned in its quest to achieve this ambitious objective.
The company’s strategy includes substantial investments in AI training chips and the construction of advanced data centers, all geared toward crafting a chatbot that rivals the sophistication of OpenAI’s GPT-4. The training process for this formidable language model is expected to commence in early 2024. Moreover, Meta’s CEO, Mark Zuckerberg, is reportedly committed to making this new AI model accessible to companies developing AI tools, thereby fostering innovation in the AI ecosystem.
A key facet of Meta’s plan involves the acquisition of additional Nvidia H100 chips for AI training. This move reflects the company’s determination to bolster its infrastructure, reducing its reliance on external cloud platforms like Microsoft’s Azure for chatbot training. This strategic shift allows Meta greater control over the development and training of its AI model.
This endeavor can be seen as a natural extension of Meta’s ongoing work in generative AI. In June, reports emerged of a chatbot for Instagram boasting a remarkable 30 personas. This development aligns with the rumors surrounding Meta’s unveiling of AI “personas,” which is anticipated to be a significant announcement later this month. These personas are expected to imbue AI interactions with human-like expressions and responses, making digital interactions more engaging and immersive.
However, Meta faces its fair share of challenges. The company has reportedly experienced turnover among its AI research teams due to the allocation of computing resources across multiple language model (LLM) projects. The competition in the generative AI landscape is fierce, with tech giants like Apple, Google, Microsoft, and Amazon actively investing in AI research and development.
Apple, for instance, is dedicating substantial resources to its AI model “Ajax,” which it claims surpasses the capabilities of even GPT-4. Google and Microsoft are integrating AI into their productivity tools, with Google eyeing the use of generative AI in Google Assistant. Amazon, too, is pursuing generative AI initiatives, potentially paving the way for a chatbot-based Alexa.
