Meta Platforms, led by Mark Zuckerberg, is postponing the launch of its new artificial intelligence model, Avocado, by approximately two months. This decision follows disappointing internal tests that revealed the model’s performance was not competitive with existing offerings from Google, OpenAI, and Anthropic.
According to reports, Meta’s AI division is contemplating a temporary license agreement with Google for its Gemini model to enhance its product capabilities. Despite efforts over the past months to improve Avocado, internal evaluations indicated that it fell short in critical areas such as reasoning, coding, and writing when compared to its rivals. Initially set for release this month, Avocado is now expected to debut around May.
Zuckerberg has significantly invested in AI development, with billions allocated for talent acquisition and infrastructure. Meta is projected to spend up to $135 billion in 2023 alone on AI initiatives, nearly double its expenditures from the previous year. While Avocado has outperformed Meta’s earlier models, it still lags behind the latest innovations from Google.
The AI efforts of Meta are closely monitored, reflecting the competitive landscape in AI development. In a memo to employees, the company announced plans to form a new AI team, fostering collaboration between various leadership figures to enhance its capabilities.
Chief AI officer Alexandr Wang has been instrumental in spearheading these initiatives, including overseeing the development of Avocado and addressing challenges that arise within the team.
Why this story matters
- The future of AI development at Meta impacts its competitiveness in the tech industry.
Key takeaway
- Meta’s decision to delay the Avocado model underscores challenges faced in keeping pace with leading AI developments.
Opposing viewpoint
- Despite setbacks, there is optimism regarding Meta’s long-term commitment to advancing AI technology.
