Big Challenges Ahead For Meta AI Chief Alexandr Wang After A Rocky First Year
One year after Meta's $14 billion acquihire of Scale AI and appointment of Alexandr Wang as chief AI officer, Meta faces scrutiny amid a 19% stock decline. Wang's Meta Superintelligence Labs has shifted away from open source models toward proprietary offerings, with mixed results. Wang now faces pressure to deliver competitive AI models and generate new revenue as Meta's infrastructure spending reaches $145 billion.
Summary
It has been one year since Mark Zuckerberg appointed Alexandr Wang as Meta's chief AI officer, following a $14 billion acquihire of Wang's company, Scale AI. During this period, Meta's shares have fallen 19%, significantly underperforming the Nasdaq, putting Wang's AI strategy under increased scrutiny.
The most significant strategic shift under Wang's Meta Superintelligence Labs has been a move away from open source models — previously offered free to developers — toward proprietary tools. The launch of Llama 4 in April 2025 was widely considered a disappointment. Wang's lab subsequently launched the Muse Spark family of models, also in April, which was seen as a step forward. However, Meta acknowledges its core models lag behind cutting-edge offerings from Anthropic and others, though they are noted for competitive energy efficiency.
In terms of monetization, Wang signaled intentions to eventually charge developers for model access, while current AI tools continue to support Meta's advertising business. Meta also introduced its first subscription offerings: Meta One Plus at $8/month and Meta One Premium at $20/month.
Wang's first year was also marked by significant personnel turbulence. He made high-profile hires from OpenAI and Anthropic, including former GitHub CEO Nat Friedman at multimillion-dollar salary levels. These large pay packages reportedly created resentment internally. The organization also lost senior employees, laid off approximately 600 workers, and faced reports of internal conflict and low morale.
Looking ahead to Wang's second year, he faces two primary challenges: delivering on his promise that upcoming Muse Spark models will be competitive with the world's leading AI models, and generating meaningful new revenue as Meta is expected to spend up to $145 billion on AI and data center infrastructure — nearly double the $72.5 billion spent the prior year. Investors are closely watching whether Meta can convert consumers, businesses, and developers into paying customers.
Key Insights
- Wang's Meta Superintelligence Labs has led a strategic pivot away from open source AI models — previously offered free to developers — toward proprietary tools, representing a fundamental change in Meta's AI philosophy.
- Meta acknowledges that its core AI models are less powerful than cutting-edge models from Anthropic and others, though the company claims competitive performance in energy efficiency.
- Wang indicated Meta eventually wants to charge developers for access to its models, marking the first time Meta has signaled intent to directly monetize its AI offerings beyond ad business support.
- High-profile hires at multimillion-dollar salary levels — including former GitHub CEO Nat Friedman — reportedly generated internal resentment, contributing to personnel losses, roughly 600 layoffs, and reports of low morale.
- Meta is expected to spend as much as $145 billion on AI and data center infrastructure in the coming year, nearly double the $72.5 billion spent the prior year, intensifying pressure on Wang to generate new revenue.
Topics
Transcript
[0:00] It's been one year since Mark Zuckerberg appointed Alex Wang, Meta's chief AI officer, as part of a $14 billion acquihire as Meta invested in Wang's company, Scale AI. With Meta shares down 19% in the past year, far underperforming the Nasdaq in that period, Meta and Wang's AI direction is under more scrutiny than ever. The biggest change in Meta's AI approach from Wang's Meta Superintelligence Labs has been a shift away from open source models, which were previously offered for free to the [0:30] developer community. But its launch of Llama 4 back in April 2025 was considered a big disappointment. Meta is now focused on proprietary tools, with Wang's lab responsible for launching the new Muse…
Full transcript available for MurmurCast members
Sign Up to AccessMore from CNBC
Watch AI Dock A Boat
CNBC's Contessa Brewer demonstrates a 36-foot Cray boat equipped with advanced marine technology, including GPS, radar, and newly launched autonomous docking capabilities. The boat's base price is $550,000, with fully equipped models featuring autopilot and autonomous docking reaching $880,000.
What Will TV Look Like In 3 Years?
Industry leaders predict that TV will continue its shift toward streaming and live content over the next three years, with cable subscribers declining further and AI enabling personalized, multilingual viewing experiences. Sports remain a critical driver of viewership as one of the few irreplaceable live experiences, while new aggregator services and platforms like Tubi and Roku Channel are expected to gain significant market share.
To keep growing, Best Buy wants to go smaller
Best Buy is launching small and medium-format stores this summer as part of a strategic pivot following years of declining sales and stock performance post-COVID. New CEO Jason Bonfig will take over from Corie Barry in October and plans to accelerate growth through this revised store strategy while capitalizing on the AI boom.
How Audi hopes to win back consumers
Audi is attempting to regain market competitiveness by launching three new premium SUVs, including the Q9 and SQ9, featuring advanced technologies like smart headlights and OLED lighting. The brand has struggled since 2020, falling behind BMW and Mercedes-Benz in quality rankings and market share, with recent losses attributed to tariffs and the end of federal EV subsidies.
Inside SK Hynix: We Went To Korea To See The World's Biggest AI Memory Buildout
SK Hynix, South Korea's memory chip giant, is undertaking a historic $720 billion expansion to triple HBM (high-bandwidth memory) capacity by 2034, driven by unprecedented AI demand. The company is building massive new facilities in Korea while establishing its first U.S. packaging fab in Indiana, though it faces traditional boom-bust cycle risks and competition from Samsung, Micron, and Chinese manufacturers.