Thank you for reading this post, don’t forget to subscribe!
Tech giants OpenAI and Meta are gearing up to release new artificial intelligence models that promise to bring machines closer to achieving superhuman cognition. Both companies have hinted at the imminent launch of upgraded versions of their large language models, including Meta’s Llama 3 and OpenAI’s anticipated GPT-5.
According to Meta’s vice-president of AI research, Joelle Pineau, the goal is to enable these models not only to communicate but also to reason, plan, and retain memory. Similarly, OpenAI’s chief operating officer, Brad Lightcap, emphasized that the next generation of GPT would address challenging tasks like reasoning, marking a significant leap in AI capabilities.
Current AI systems excel at specific tasks but lack the ability to handle complex sequences or retain information over time, leading to what Meta’s chief AI scientist, Yann LeCun, described as “stupid mistakes.” By integrating reasoning capabilities, AI models can search for answers, plan actions, and predict outcomes, paving the way for artificial general intelligence.
Meta plans to integrate its new AI model into platforms like WhatsApp and its Ray-Ban smart glasses, envisioning scenarios where AI assistants guide users through tasks like troubleshooting a broken coffee machine. Similarly, OpenAI aims to enhance GPT’s capability to tackle longer and more complex tasks, driven by advancements in reasoning abilities.
As AI technology continues to evolve, Meta and OpenAI’s efforts represent a significant milestone towards achieving human-level cognition in machines, ushering in an era where AI systems play an integral role in everyday interactions and tasks.
