Meta Platforms Set to Launch Upgraded AI Model, Llama 3, in July

Meta Platforms is gearing up to unveil an upgraded version of its artificial intelligence model, Llama 3, in July, according to a report by The Information on Wednesday.

The focus of Meta researchers is to refine Llama 3 so that it can offer more nuanced responses to sensitive or contentious queries posed by users.

The aim is to imbue the model with the capability to provide contextual understanding to questions it deems controversial.

This development effort seeks to enhance the utility of Meta's latest Language Learning Model (LLM) amid growing competition in the AI space.

Notably, Meta's move comes in the wake of rival Google's decision to temporarily halt the image-generation feature on its Gemini AI.

This pause followed instances where the AI produced historically inaccurate images.

Initial tests conducted by The Information revealed that Meta's previous iteration, Llama 2, which powers chatbots across its social media platforms, was hesitant to respond to inquiries deemed less controversial.

For instance, it would refrain from answering questions related to pranks, warfare tactics, or vehicle tampering.

However, with Llama 3, Meta aims to bridge this gap in comprehension.

The new model is anticipated to decipher queries such as "how to kill a vehicle's engine" with greater precision.

This clarification is essential as the query may pertain to simply shutting off the engine rather than causing physical harm to the vehicle. The report draws from multiple sources familiar with Meta's development endeavors.

Meta's upcoming release of Llama 3 underscores its commitment to advancing AI technology to better cater to user needs and expectations in an ever-evolving digital landscape.

Source: Priyanka.G and Yuvraj Malik / Reuters