Google's Efforts in Developing Human-like Reasoning AI to Compete with OpenAI

TapTechNews October 3rd news, Bloomberg reported yesterday (October 2nd) that Google is developing a human-like reasoning AI and hopes to catch up with OpenAI's o1 model in terms of progress.

OpenAI has recently introduced the o1-preview and o1-mini models for ChatGPT, which can simulate the human thinking pattern and achieve general reasoning, enabling the AI to think.

TapTechNews quoted Bloomberg's report that Google is also actively promoting the landing of the general reasoning AI model at present. This AI model is better at handling mathematics and programming, but consumes more processing power and energy.

Insiders said that like OpenAI, Google is also trying to use a technology called chain of thought prompting to approach human reasoning ability. In this technology pioneered by Google, the software will pause for a few seconds and then respond to written prompts. At the same time, in the background and invisible to users, it will consider many related prompts and then summarize what seems to be the best answer.

The new model uses the step-by-step reasoning prompting technology and can consider multiple responses and self-debate to find the correct solution.

Bloomberg said that Google is not far behind in the general reasoning AI model and already has relatively sufficient technical reserves. The article did not mention the release time, only revealing that the AI model developed by Google will have new progress in the next few months.