Meta's No Language Left Behind model announcement caused investors to withdraw funding from African language startups, forcing organizations working on 55 African languages to shut down. Investors told the startups Meta had "solved" translation, making their specialized work obsolete, AI ethics researcher Timnit Gebru reported.
OpenAI representatives have delivered ultimatums to small language AI organizations across multiple markets: collaborate for minimal payment or face closure. "OpenAI is going to put you out of business soon because we're going to make our models better in your language," representatives told organizations, according to Gebru. The company offered to purchase data for what sources called "peanuts."
This pattern repeats globally. When any Big Tech company announces models covering specific languages, investors pressure smaller organizations in those markets to close. The dynamic blocks purpose-built AI solutions designed for specific communities and languages from securing funding.
AI ethics researchers Gebru and Abeba Birhane argue resource-intensive giant models pose safety risks. Models designed for undefined tasks lack clear safety parameters, they contend. The development process relies on massive data collection critics call appropriation without consent.
Birhane identified "AI for good" rhetoric as deflection. Companies cite social benefits to counter resistance movements, she said. "AI for good allows companies to say 'Look, we're doing something good! Everything about AI is not bad. And you can't criticize us,'" Birhane stated.
The researchers challenge claims that giant models benefit marginalized communities. Gebru describes the development as "stealing data, killing the environment, exploiting labor." Training massive models demands energy consumption that raises sustainability concerns.
The tension between centralized giant models and specialized solutions represents a fundamental divide in AI development. Purpose-built models for specific languages offer targeted benefits without resource demands of general-purpose systems. Big Tech's market dominance prevents these alternatives from securing funding.
Small language AI displacement particularly affects Global South communities, where specialized models address linguistic and cultural needs general-purpose systems may not serve. The consolidation threatens diversity in AI development as Western tech giants expand language coverage using approaches critics question.