Meta's AI is capable of translating numerous under-resourced languages
The initiative involves over 200 languages
An AI system developed by Meta aims to support endangered languages by scaling up machine translation for languages with limited digital resources.
Here are a few key points:
The initiative, detailed in Nature, involves translating over 200 languages, including Tswana, Dari, and Samoan.
This effort seeks to bridge the digital divide and help preserve endangered languages by bringing them online.
However, the success of these AI models hinges on high-quality human-generated data.
Researchers and specialists must collaborate closely with native-speaking communities to ensure accuracy and cultural integrity in translations.
Neglecting this human aspect risks poor data quality and could inadvertently harm the very languages the technology aims to support.
“The researchers, based at Meta AI, Meta’s research division in New York City, the University of California, Berkeley, and Johns Hopkins University in Baltimore, Maryland, set out to expand the number of low-resource languages that their model translates as part of Meta AI’s ‘No Language Left Behind’ programme. They selected languages that were present in Wikipedia articles, but had fewer than 1 million sentences of example translations available online.”
Things I’m reading today
LinkedIn halts group-based ad targeting amid European regulatory concerns (link)
LinkedIn has announced that it will no longer allow advertisers to target users based on data from LinkedIn Groups. This decision follows a complaint from civil society groups to the European Commission (EC) about potential violations of the Digital Services Act (DSA), which governs online content and advertising practices.
The EC sought clarification from LinkedIn regarding the use of sensitive personal data for targeted ads. Although LinkedIn contends it complies with the DSA, it has decided to remove this targeting option to avoid any misconceptions.
Adobe's updated terms spark outrage among pro users over content rights concerns (link)
Adobe sparked controversy among its professional users by updating its terms of service, which included alarming language suggesting Adobe might claim rights over users' work. Users had to agree to these terms to continue using the apps, request support, or uninstall them, leading to widespread anger and confusion. Initially, Adobe issued a dismissive statement, claiming the terms had been in place for years and were merely clarified.
However, the company later provided a detailed blog post explaining that the terms were meant to be transparent about content checks, including manual and automated moderation for generating thumbnails and scanning for prohibited content like CSAM and spam.
Adobe assured users that it does not train its Firefly Gen AI models on customer content and that it does not assume ownership of user work. Adobe promised to clarify the terms of use acceptance process in response to the backlash.
