Microsoft-backed OpenAI and Meta will soon release artificial intelligence (AI) models capable of reasoning and planning, the Financial Times (FT) reports.
Meta has announced plans to release Llama 3 in the coming weeks, and OpenAI has revealed that its next AI model, GPT-5, will also be introduced soon, the report added.
Click here to follow our WhatsApp channel
As reported by the FT, Joelle Pineau, Meta's vice president of AI research, said the company is actively exploring the capabilities of large-scale language models that involve reasoning, planning, and memory functions. These models, which are the backbone of popular generative AI applications such as ChatGPT and Llama, are being studied for further enhancements, he said.
OpenAI Chief Operating Officer Brad Lightcap said the next version of GPT is aimed at tackling difficult tasks such as inference, the FT reported.
Lightcap said current AI systems have limited capabilities and have only scratched the surface. But AI will evolve to handle more complex tasks in sophisticated ways, he said.
OpenAI launches GPT-4 Turbo
On April 10th, OpenAI launched an upgraded version of GPT-4 Turbo. It is currently available via an application programming interface (API) and can be accessed by ChatGPT Plus subscribers. This latest version, his GPT-4 Turbo, represents the latest advances in the company's generative AI models, trained on data through December 2023.
According to a report in Indian Express, the revised model has enhanced data processing capabilities that enable text analysis in JSON format and execution of function calls that generate JSON code snippets. These snippets help developers automate tasks within integrated applications, such as sending emails, posting online content, and completing purchases.
The upgraded GPT-4 Turbo can generate fully functional websites by generating authentic code and interpreting image content and context to produce meaningful output.
