Meta's $14B AI gamble with Alexandr Wang falls short as developers turn skeptical

Meta AI Faces Mounting Pressure After Massive Investment

One year after Meta poured over $14 billion into acquiring roughly half of Scale AI and bringing Alexandr Wang along with his top engineers onboard, the social media giant finds itself struggling to convince Wall Street and developers that it can compete in the AI arena1

1

. The company's stock has declined 18% over the past 12 months, making it the worst performer among tech megacaps alongside Microsoft, even as Meta reported 33% revenue growth in the first quarter—its fastest expansion rate since 20211

1

. Mark Zuckerberg now faces the critical challenge of transforming Wang's technical achievements into tangible financial results, as investor patience wears thin over the company's ability to demonstrate AI monetization beyond its core advertising business.

The Shift From Open-Source to Proprietary AI Model

Meta's AI strategy underwent a dramatic transformation following the disappointing reception of Llama 4 in April of last year, which failed to captivate developers and exposed vulnerabilities in the company's open-source approach1

1

. The Llama models, which Meta had positioned as freely accessible alternatives to paid offerings from competitors, ultimately became what industry experts now call a strategic blunder1

1

. This setback prompted Zuckerberg's shocking $14.3 billion investment in Scale AI just two months later, establishing Meta Superintelligence Labs under Wang's leadership to pivot toward Meta's proprietary AI model development1

1

. The resulting Muse Spark model, delivered in April this year, marks Meta's first jump into proprietary foundation models and represents a fundamental departure from the company's previous commitment to open-weight AI1

1

.

Developer Adoption of AI Remains Lukewarm

Unlike its open-source predecessors, Muse Spark was designed specifically for internal integration across Meta's ecosystem, including Facebook, Instagram, Ray-Ban Meta glasses, and the standalone Meta AI app1

1

2

2

. However, this inward focus has created significant challenges in building trust among the broader developer community3

3

. Rob May, CEO of startup Neurometric, noted that "the AI community largely ignores Meta at this point," characterizing Muse Spark as a "yawn" since the technology remains largely inaccessible to external developers1

1

. Limited access has slowed adoption, with Muse Spark mainly working inside Meta's own applications while outside developers have very little access3

3

. Safety concerns also influenced this decision, as Wang confirmed that Muse Spark showed risk signals during testing, particularly in sensitive areas, leading Meta to keep the model private for better control3

3

.

Challenges in Monetizing AI Tools Beyond Advertising

Ralph Schackart, an analyst at William Blair, emphasized that "investors are looking for Meta to monetize a new AI-first product, beyond the substantial positive impact AI is having on enhancing the advertising models"1

1

2

2

. Meta still counts on its advertising business for 98% of revenue, and historical attempts to diversify have proven unsuccessful1

1

. The company has started testing AI subscription plans as part of efforts to expand beyond online ads, but clear returns have not yet materialized1

1

3

3

. Meta's AI investments face additional scrutiny following the layoff of 8,000 workers last month, with sources indicating tension at the top of the AI organization as Wang and other high-profile hires face severe pressure to justify the company's massive spending2

2

.

Meta's AI Strategy Against Fierce Competition

Thomas Randall, an analyst at Info-Tech Research Group, acknowledged that while Meta hasn't taken the "most optimized route," the company would be "lost" without Zuckerberg's wallet-opening for Wang and other big-name AI hires in what he termed a "strategic rebuild"1

1

2

2

. Meta remains far behind OpenAI, Anthropic, and Google in the AI market, with firms like Anthropic's Claude gaining attention by offering reliable and widely used tools1

1

3

3

. AI competitiveness demands that Meta match this trust while improving performance, particularly as the company reportedly postponed plans to share its latest AI models with developers with no schedule for release2

2

. Schackart wants to see "tangible evidence of a growing list of new, AI-first products created by Muse Spark, even if monetization lags," which he says is what investors are watching for1

1

. Despite protecting a $200 billion-a-year business, Meta must demonstrate it can attract paying users for its AI tools rather than solely using the technology to enhance its existing operations1

1

. Andrew Moore, CEO of enterprise startup Lovelace and former Google Cloud AI chief, suggested it's not too late for Meta to find its lane in the increasingly crowded AI landscape1

1

.