”
Nick Clegg, Meta’s President of Global Affairs, has expressed that current Artificial Intelligence (AI) models are still “quite stupid,” downplaying the risks associated with the technology.
He believes the hype surrounding AI has outpaced its current capabilities, stating that existing models are far from reaching the stage where AI achieves autonomy and independent thinking. Meta recently announced that its large language model, Llama 2, will be available as open source, allowing widespread usage for commercial businesses and researchers.
Llama 2, which powers chatbots like ChatGPT, operates by connecting data points in massive text datasets and predicting the next word in a sequence. However, Clegg asserts that the existential threat warnings issued by some AI experts pertain to systems that do not yet exist. Despite Meta’s move to open-source Llama 2, the decision has sparked division within the tech community, considering the potential risks associated with such a powerful tool.
The open-source approach allows Meta to gather valuable user testing data, identifying bugs, issues, and improvements. However, concerns linger over the tool’s potential misuse, as previous iterations of chatbots have been manipulated to spread hate speech and misinformation. Meta’s partnership with Microsoft to make Llama 2 available via Azure adds another dimension to the development. As Microsoft has heavily invested in OpenAI, the creator of ChatGPT, industry watchers wonder whether the AI landscape could become dominated by a few major players, potentially hindering healthy competition in this evolving sector.
