Maya PH’s open-source LLM, Godzilla 2, surpasses ChatGPT in truthfulness
Maya Philippines released Godzilla 2, an open-source large language model that outperformed major proprietary systems like ChatGPT and GPT-4 in truthfulness metrics. This achievement highlights the growing capability of open models to compete with commercial giants, demonstrating that transparency does not require sacrificing accuracy or reliability in critical AI evaluations. The release underscores the importance of minimizing misinformation, as accurate AI becomes increasingly vital for public trust and search engine integration. By prioritizing truthful responses, developers address the urgent need for regulation and ethical standards, ensuring that AI aids rather than hinders societal navigation of an increasingly complex information landscape. This initiative is highly relevant to open data as it advocates for collaborative innovation over corporate monopolies. By making the model accessible, Maya aims to improve AI literacy in the Philippines and encourages broader data sharing, proving that open ecosystems can drive technological progress and democratize access to advanced artificial intelligence tools.
Source: rappler.comPublished on 2023-09-07
Related news
- Technology Innovation Institute Introduces World’s Most Powerful Open LLM: Falcon 180B
- OpenAI and Microsoft accused of stealing data to train ChatGPT in new class-action suit
- The Guardian, New York Times, CNN figure in growing list of sites blocking OpenAI crawler
- Right to Information Month launched to increase awareness
- ¿Cómo saber si estafadores lo tienen en la mira?