Ian Goodfellow, Director of machine learning within Apple, would have decided to leave the firm. The journalist from The Verge Zoë Schiffer reports that her decision would be motivated by the giant’s return to work policy.
Apple employees have demonstrated during the COVID-19 health crisis that both productivity and efficiency can be achieved even while remote. It was therefore not surprising that employees asked to work from home after Apple announced the return to face-to-face last September.
During the health crisis, Apple employees worked remotely. They have acquired new habits and have been able to demonstrate efficiency in their missions. With the gradual return to face-to-face, employees returned to their offices in September 2021. However, this new policy of ending teleworking was not well received internally.
This project to return to office life was postponed to December 2021 due to the resurgence of COVID-19 cases and was finally adopted as a hybrid last month, also offering the possibility of working remotely under certain conditions. Other large companies have adopted a hybrid work policy like Meta or Airbnb. The social networking giant allows half of its employees to work remotely while Airbnb now lets its employees work from anywhere in the world. Within Apple, the demand for more flexible work policies continues to grow.
Ian Goodfellow, a great figure in the machine learning, has been with Apple for more than 3 years, after leaving Google to join the Apple firm in April 2019. He is known as the inventor of generative adversarial networks, he was also the most experienced ML expert quoted from Apple. Opposed to the current work policy, he wrote to staff: “ I firmly believe that more flexibility would have been the best policy for my team “.
He would thus have made the decision to leave his post. A departure that would be a blow to Apple, especially given how the machine learning has become an integral part of the company’s products including Siri, Face ID and of course the M1 and M1 Ultra chips, presented a few months ago, which have a 16-core CPU and 20 cores for machine learning.