It is speculated that Google is developing a 'thinking' artificial intelligence to compete with OpenAI.
Artificial intelligence more similar to humans.
Google's artificial intelligence might soon receive an update that grants it a "reflection" capability. Recently, OpenAI incorporated new models o1-preview and o1-mini into ChatGPT, designed to "think a little more before responding." Reports suggest that Google is also making progress in the development of an AI system that more effectively mimics the human ability to reason, likely intended for its Gemini platform. So far, the company has not issued an official comment on the matter.
These more advanced models are said to be more effective at solving complex problems that require multiple steps, covering areas such as mathematics and programming, by not rushing their responses. This apparently implies a higher consumption of energy and processing resources, as OpenAI has indicated regarding the o1 models.
Although Bloomberg has not specified a release date for Google's "thinking" AI, it has been noted that there have been advancements in recent months, easing concerns within the company about a potential lag behind OpenAI. There is speculation that these innovations might be available sooner than expected.
The focus of these new models is based on a method known as "chain-of-thought prompting," where the AI evaluates different steps and responses, considering various options and deliberating on the correct solution. With ChatGPT and the o1 models, users can visually observe a message indicating that the AI is in the process of "thinking." Preliminary tests suggest that these models are more competent than GPT-4o at solving puzzles and tackling complex problems. It is presumed that Google will take a similar route with its own technology.
For now, OpenAI has restricted access to the o1-preview and o1-mini models to paid subscribers of ChatGPT. This raises the question of whether Google will implement a similar requirement for access to the equivalent of Gemini, which currently has a monthly cost.
Google's AI development plans are constrained by ethical concerns, the need to maintain a trustworthy image, and internal competition within the company. Meanwhile, OpenAI, Meta, and Microsoft continue to make strides in their own artificial intelligence updates, adding pressure on Google to advance its initiatives.
