The AI race is heating up as a perceived laggard makes a major push. Just one day after SpaceX's AI division unveiled a new large language model, Meta Platforms, Inc. launched its latest model, Muse Spark 1.1, on Thursday. Similar to SpaceX, Meta claims this new model's performance is on par with flagship products from leading labs like OpenAI, Anthropic, and Alphabet. However, it will take a few days to see if developers agree. (This launch coincides with OpenAI's release of its latest-generation model, GPT 5.6).
Aggressive Pricing Strategy
The most significant signal from this launch is Meta's plan to invest heavily in subsidizing model usage costs, aiming to seize market share from rivals. In an interview with Bloomberg, Meta CEO Mark Zuckerberg stated the company's model pricing would be "extremely competitive and attractive." Reports indicate Meta's pricing is about a quarter of the public rates for top models from Anthropic and OpenAI, effectively engaging in a price war.
This pricing strategy makes profitability from AI services impossible in the short term, but that is clearly not Meta's current goal. The company understands that AI compute costs are a major pain point for enterprise clients, who are actively seeking ways to control expenses by either using open-source models or offloading some AI tasks to older, cheaper models. In this environment, Meta's playbook is to attract customers with rock-bottom prices to try its new model, then retain them. Once it gains a foothold in the market, it can raise prices later. This tactic mirrors strategies used by companies in other sectors, like streaming service Netflix or ride-hailing giant Uber. The strategy might work, but competition in the industry is fierce.
Sustained and Substantial Investment
Meanwhile, Meta's investment in AI will continue to surge. The company had previously planned capital expenditures of $125 to $145 billion this year for AI data centers. Beyond that, operating expenses for external cloud services and AI talent are also set to rise significantly, driven by the AI business. The question remains: how much is Zuckerberg willing to spend to secure a leading position in the AI race?
Meta's heavy bet on this new frontier is not surprising. Since 2020, its Reality Labs division has accumulated losses of $87 billion, primarily due to Zuckerberg's metaverse strategy. Despite the massive investment, Meta has been scaling back its metaverse ambitions over the past two years to pivot fully toward AI.
Market Reaction and Leadership Change
Capital markets have reacted positively to Meta's AI plans. Although the stock has shown overall weakness this year, its price rose 4.7% on Thursday.
In a separate development, Coinbase announced on Thursday that Paul Grewal, its Chief Legal Officer of nearly six years, is stepping down. Grewal posted on social media platform X, using the common departure phrase: "It's time for a new chapter."
During his tenure at Coinbase, Grewal handled several major events: he led the company's defense against a 2023 lawsuit by the U.S. Securities and Exchange Commission, which accused Coinbase of operating an unregistered securities exchange and brokerage. The SEC dropped the lawsuit last year after a new administration took office. Recently, he spearheaded negotiations for the cryptocurrency market regulation bill known as the FIT Act, which remains stalled in Congress. The crypto industry hopes to advance the bill, which would establish a comprehensive regulatory framework for tokens and trading platforms, before the midterm elections. Coinbase stated that Grewal will serve as an advisor to the company until the end of October. Speculation is rife about his next move, with a high probability it will be at an AI company.
Comments