Mark Zuckerberg is not shying away from big bets in the artificial intelligence arms race.
Meta recently made a move to acquire Perplexity AI, but the negotiations did not end in a deal, according to sources familiar with the matter who wished to remain unnamed due to the confidentiality of the discussions.
Accounts differ on why the talks collapsed. One person characterized it as a mutual decision to step away, while another indicated Perplexity opted to leave the table. Perplexity has declined to provide any comment, while Meta has yet to respond to requests for their side of the story.
The failed acquisition hints at Meta’s sheer determination to match the pace set by OpenAI and Alphabet. The social media titan has invested $14.3 billion in Scale AI, securing nearly half of the startup’s equity, although without voting rights. This massive financial commitment marks one of the most significant moves yet as Meta races to strengthen its AI capabilities.
Meta’s Expanding AI Ambitions
Following the investment, Scale AI’s founder Alexandr Wang, together with a select group of Scale employees, will transition to Meta as part of the agreement. This move folds them into Meta’s growing AI leadership, which has been swelling with high-profile names in recent months.
Earlier this year, Meta made a separate approach to Safe Superintelligence, a company that, according to reports, reached a valuation of $32 billion during an April fundraising round. The company’s CEO Daniel Gross and former GitHub CEO Nat Friedman are also joining Meta, where they will report to Wang in building new AI products.
The AI competition between tech’s biggest players often spills into talent wars. OpenAI CEO Sam Altman acknowledged on his brother’s “Uncapped” podcast that Meta had offered some OpenAI staff signing bonuses up to $100 million, topped by even larger annual compensation. Altman put it bluntly: “I’ve heard that Meta thinks of us as their biggest competitor. Their current AI efforts have not worked as well as they have hoped and I respect being aggressive and continuing to try new things.”
All eyes, it seems, are now fixed on Meta’s next move, as the company’s continued push toward [open foundation language models](https://ai.meta.com/research/publications/llama-2-open-foundation-and-fine-tuned-chat-models/ “open foundation language models” target=blank) and [AI leadership changes at Meta](https://critiqs.ai/ai-news/meta-faces-ai-leadership-crisis-as-llama-team-departs/) set the stage for the industry’s next phase.
