Tech Giants Set to Launch Enhanced AI Models Amidst Intensifying Competition in the Industry
In a bid to push the boundaries of artificial intelligence (AI) capabilities, tech behemoths Meta and OpenAI are gearing up to unveil upgraded AI models capable of advanced reasoning and planning functions, according to a recent report by the Financial Times (FT).
Meta has announced plans to release Llama 3 in the coming weeks, while OpenAI is preparing to introduce its latest AI model, ‘GPT-5’, in the near future, as outlined in the FT report.
Joelle Pineau, vice-president of AI research at Meta, highlighted the company’s exploration into leveraging large language models for tasks involving reasoning, planning, and memory functions. These models, which underpin popular generative AI applications like ChatGPT and Llama, are undergoing further enhancements to unlock their full potential.
Brad Lightcap, chief operating officer (COO) of OpenAI, emphasized the upcoming version of GPT’s focus on addressing challenging tasks, including reasoning. He acknowledged the current limitations of AI systems but expressed confidence in their evolution towards handling more intricate tasks in a sophisticated manner.
The competition in the AI arena is heating up, with rival companies such as Google and Anthropic gearing up for their own AI launches. Meta plans to integrate its latest AI model into platforms like WhatsApp and Ray-Ban smart glasses, while also diversifying Llama 3 for various applications and devices in the coming months.
In related news, OpenAI recently rolled out an upgraded edition of GPT-4 Turbo on April 10, now accessible via an application programming interface (API) for subscribers of ChatGPT Plus. This latest iteration, GPT-4 Turbo, represents OpenAI’s cutting-edge advancements in generative AI models, incorporating data up to December 2023.
According to reports from the Indian Express, GPT-4 Turbo boasts enhanced data processing capabilities, enabling text analysis in JSON format and the execution of function calls to generate JSON code snippets. These snippets facilitate task automation within integrated applications, ranging from email communication to content creation and online transactions.
Furthermore, the upgraded GPT-4 Turbo demonstrates the ability to generate fully functional websites by crafting authentic code and interpreting image content and context to deliver meaningful outputs.
