Rio de Janeiro's claimed locally-developed large language model appears to be constructed from existing models rather than built from scratch, according to findings shared on GitHub.
Developers investigating the Nex-N2 project discovered evidence suggesting Rio's touted 'homegrown' LLM is actually a merge of pre-existing language models, raising questions about the authenticity of the claims.
The discovery was documented in a GitHub issue that gained traction across tech communities, attracting 149 upvotes and 83 comments on Hacker News. The findings indicate the model combines components from established LLMs rather than representing original development work.
This development highlights ongoing scrutiny in the AI community around claims of local or proprietary model development. As organizations worldwide race to build AI capabilities, the distinction between genuine innovation and repackaging existing work has become increasingly important for transparency and credibility.
The implications extend beyond Rio, reflecting broader concerns about how AI development is marketed and the standards for claiming original model creation in an era of rapid LLM proliferation.
Trust—not technology—represents the primary obstacle to widespread AI adoption, according to Nilam Ganenthiran, CEO of Beacon and former Instacart President. Safety concerns are actively suppressing economic growth and innovation.
Google is testing a new Gemini feature that allows the AI to make phone calls to businesses on behalf of users. The capability will initially roll out to Pixel 11 owners in the U.S. with paid Gemini subscriptions.
Artificial intelligence is achieving benchmark performance levels at rapidly declining costs, dropping roughly 13x annually. The decline stems from both hardware improvements and algorithmic progress, which MIT estimates at 3x per year.
Venture capitalist Vinod Khosla said the tech industry has damaged itself through poor communication about AI, while arguing that AI development should continue at its current pace.