Nick Clegg defends release of open-source AI model by Meta
Nick Clegg, Meta’s president of global affairs, defends the open-sourcing of its Llama 2 model by arguing that current fears regarding AI dangers are largely fueled by hype rather than reality. He asserts that existing models are nowhere near the hypothetical "super-intelligent" levels often cited in existential risk warnings. By characterizing current technology as relatively limited, Clegg suggests that releasing these tools freely is a responsible step, contrasting the actual capabilities with the exaggerated scenarios driving public anxiety. This perspective directly impacts the open_data ecosystem by challenging the notion that open access inherently leads to misuse. While some experts compare open-sourcing large language models to distributing dangerous weapons, Meta contends that broad community access enhances safety through collective scrutiny. Clegg argues that keeping such powerful models within the exclusive control of tech giants is riskier than allowing a wider range of researchers and developers to inspect, adapt, and improve them, thereby democratizing knowledge rather than concentrating it. Ultimately, the article highlights the growing importance of open-source AI in shifting the competitive landscape away from closed, proprietary systems. This move underscores a critical trend where transparency and shared data resources are becoming central to technological advancement and safety. For the open_data community, this validates the strategic value of accessible models, suggesting that widespread availability can drive innovation and accountability more effectively than restrictive corporate gatekeeping, despite ongoing debates about self-regulation.
Source: theguardian.comPublished on 2023-07-20