In a week marked by significant advancements in artificial intelligence, both Meta and OpenAI have unveiled their latest AI models, promising to reshape the landscape of AI technology and its applications.
Meta’s Llama 3.1: A New Frontier in Open AI
On 23th of July, Meta announced the public release of Llama 3.1 405B, touted as “the world’s largest and most capable openly available foundation model.” This model, alongside its smaller versions, Llama 3.1 70B and 8B, represents a major leap in AI capabilities. Meta claims Llama 3.1 offers state-of-the-art performance in general knowledge, steerability, mathematics, tool use, and multilingual translation
Meta CEO Mark Zuckerberg emphasized the importance of open-sourcing AI technology, likening Llama to the Linux operating system, which revolutionized computing through its open-source model. Despite the open-source branding, Meta imposes certain restrictions on Llama 3.1’s commercial use.
OpenAI’s SearchGPT: Challenging Google’s Dominance
Not to be outdone, OpenAI introduced SearchGPT on 25th of July, a prototype online search tool powered by a large language model. This launch marks OpenAI’s direct challenge to Google’s supremacy in the search engine market. Currently in beta testing with selected news publishers, SearchGPT aims to deliver real-time, conversational search results, incorporating citations and links to relevant sources similar to Perplexity’s AI-powered search platform. OpenAI positions SearchGPT as a valuable tool for publishers, offering up-to-date information directly from the web through conversational AI.
By directly integrating real-time web data into conversational AI responses, there is a risk of amplifying inaccuracies and biases present in the source data. Moreover, the privacy of users querying the AI is a critical concern, as detailed queries can reveal sensitive personal information. OpenAI must implement stringent data protection measures and ensure transparency in how data is collected, processed, and stored.