Google is working to create an artificial intelligence model that can reason like a human. Due to this, the IT giant intends to increase competition with OpenAI, which has already presented a similar product called o1. Bloomberg writes about this with reference to its own informed sources.
Image source: geralt/Pixabay
The report said that in recent months, Google has made significant progress in developing artificial intelligence models that can handle multi-step problems in areas such as mathematics and programming. According to the source, like OpenAI, Google is trying to create a reasoning AI algorithm using a technique called “chain of thought.” It implies that before giving an answer to a written request, the AI algorithm pauses for a few seconds to analyze topics related to the request and then give a generalized answer to the question posed. Google officials declined to comment on the matter.
Google and OpenAI are locked in a tight battle for dominance in artificial intelligence. The standoff has intensified with the advent of the AI bot ChatGPT, which some investors believe could eventually eliminate the need for Google’s search engine. The IT giant is trying in every possible way to avoid this, for which the company’s leading research development teams were united into the DeepMind division. Despite this, Google continues to move more slowly when it comes to launching new AI products, pausing to consider ethical issues, the need to meet public expectations of the company’s brand credibility, and more.
An anonymous source for the publication said that after OpenAI introduced the o1 algorithm in mid-September, some DeepMind employees were disappointed that Google was lagging behind in the race for leadership in the field of AI. However, employees are no longer as concerned as they were after the launch of ChatGPT, since Google has now released some of its own developments.