fbpx

Google Sparrow

google sparrow - 226lab agence créative numérique

We have already talked about Chat GPT and how it is rapidly gaining popularity in the world of technology. This innovative AI has amazed everyone with its human-like writing ability, sparking a wave of interest in conversational chatbots and artificial intelligence, especially regarding its impact on the world of work. This innovation is still in its early stages, yet competition is already beginning to emerge. Google, in fact, does not want to be left behind, and to stay in the game, it has announced the creation of Google Sparrow.

How Google Sparrow Came to Be

Temps interviewed Demis Hassabis, the head of DeepMind, a subsidiary of Alphabet, Google’s parent company, and an AI lab.

DeepMind first came into contact with Google in 2013 when it presented a prototype of artificial intelligence trained to play a video game, based on apprentissage par renforcement. It was soon after acquired by Google.

After the acquisition, DeepMind designed the famous algorithm AlphaGo, which, through reinforcement learning, managed to defeat a Go champion multiple times.

While Google continues private AI research, another project, OpenAI, is initiated and opened to the public in 2022 with the ChatGPT chatbot.

Under pressure from competition, DeepMind is considering releasing its own chatbot, Google Sparrow, as announced by Hassabis to The Times. There are no definite release dates yet, but a “private beta” is likely to be launched in 2023.

Google Sparrow Features

Google Sparrow will use apprentissage par renforcement, learning from its mistakes. In comparison to Chat GPT, it will also cite the sources it uses and provide more accurate responses.

 

Concerns about AI

Hassabis expressed concern in his interview with Time about the pace of things: the faster AIs are launched, the more likely there are to be negative effects on society. Artificial intelligence will be one of the most powerful technologies ever, and careful attention must be paid to its development to avoid unpleasant consequences. It’s better to have safe and ethical progress than overwhelming and harmful progress.

What are your thoughts on this?