AI translation disrupts language learning: commodifying culture while obscuring colonial legacies of linguistic erasure
Original framing: “If AI can translate instantly, why learn another language?” — Phys.org
By 2035, AI translation could render 80% of global languages non-viable for digital interaction, accelerating linguistic extinction at a rate unseen since the Bronze Age collapse. Scenario modeling suggests that platform-dependent translation will create a 'digital linguistic underclass' of speakers of low-resource languages, while reinforcing English as the default lingua franca of the internet. Alternative futures include decentralized, community-owned translation models and the revival of endangered languages through AI-assisted preservation.
The AI translation boom is not merely a technological leap but a continuation of colonial linguistic erasure, where Silicon Valley’s extractive logics replicate historical patterns of suppression under the guise of progress.
Indigenous knowledge systems, which treat language as a living, relational force, are systematically devalued by platforms that reduce communication to data packets, while marginalized communities—Deaf signers, refugees, and indigenous speakers—are rendered invisible in the design process. The future modeled by unchecked AI translation is one of linguistic monoculture, where the richness of human expression is flattened into a corporate-controlled lingua franca. Yet alternatives exist: from Māori-led cooperatives to open-source models like Masakhane, the path forward requires dismantling the power structures that treat language as a commodity. The trickster’s lesson is clear—what appears as liberation (instant translation) is often another form of enclosure, and the true work lies in reclaiming language as a site of communal sovereignty and cultural resilience.