Apple continues its AI shopping with a promising French start-up specializing in quantification


Samir Rahmoune

April 23, 2024 at 7:53 a.m.

0

The Apple logo, displayed on the window of an App Store © Alexandre Boero / Clubic

The Apple logo, displayed on the window of an App Store © Alexandre Boero / Clubic

The American giant has just made a new acquisition. And it’s a French start-up working on algorithm issues which is joining the ranks of Apple!

The firm founded by Steve Jobs has a lot of money, which allows it, if necessary, to invest a lot and very quickly. A very useful facility at the moment for the group, which must catch up in the field of artificial intelligence in the face of Gemini and other ChatGPT. To do this, it can either recruit more people internally to work on the subject, or simply acquire companies in the field. What has just been done with the acquisition of the Parisian start-up Datakalab!

Apple acquires Datakalab

Apple is active in the field of artificial intelligence. New proof is brought to us by this information from the magazine Challenges, thanks to whom we learn that the American giant has bought 100% of the shares of the French start-up Datakalab. The operation, completed on December 17, and notified to the European Commission, could represent an investment of several million euros. Even if no figures have been communicated by Apple.

Datakalab is a company founded in Paris in 2017, which originally worked on the use of AI to read emotions on the faces of human beings. A field still too closely connoted as close to facial recognition, which pushed the company to reorient itself in 2022 towards the field of algorithm compression, via what is called “quantification. »

AI for the iPhone in the crosshairs?  © Tada Images / Shutterstock

AI for the iPhone in the crosshairs? © Tada Images / Shutterstock

Quantification, the future of AI?

This method aims to reduce the complexity of calculations when performing a task, by reducing the precision of the data. This would make it possible to reduce the resources allocated to this work, particularly in terms of chip power. “ With compression, we can go from an Nvidia semi-processor for 10,000 euros to an STMicroelectronics chip for a few cents » summarizes to the media Challenges a person close to Datakalab.

Suffice it to say that this is a method that perfectly meets Apple’s current needs. The American group wants to develop AI systems that can operate locally, without being connected to the cloud, with the aim of integrating them into the next iPhones. However, without a connection, it is the device that must do all the work, unlike what is currently happening with large chatbots. Quantification appears in this context to be extremely attractive for Apple!

Source : Challenges, 01net

Samir Rahmoune

Tech journalist, specializing in the impact of high technologies on international relations. I am passionate about all the new developments in the field (Blockchain, AI, quantum...), the...

Read other articles

Tech journalist, specializing in the impact of high technologies on international relations. I am passionate about all the new developments in the field (Blockchain, AI, quantum...), energy issues, and astronomy. Often one foot in Asia, and always ready to put on the gloves.

Read other articles



Source link -99