Search:

Link

Third Generation AirPods Likely to be 'Unleashed' at Apple Event

Charlotte Henry ·

As I, and many others, suspected, it seems we will finally get the third generation of AirPods at the ‘Unleashed’ event on Monday. Leaks from Chinese social media site Weibo, picked up by MacRumors, suggest the earbuds will have an updated design.

Now, as Apple has confirmed its likely last event of the year for this coming Monday, October 18, all the indications suggest that new ‌AirPods‌ will also be announced alongside redesigned MacBook Pros. Weibo leaker @PandaIsBald, which accurately reported the launch of the baseline ninth-generation iPad for Apple’s last event, has claimed that alongside “M1X” Macs, the third-generation ‌AirPods‌ with an updated design will also be announced…The refreshed ‌AirPods‌ are expected to take design cues from the ‌AirPods Pro‌ and leaked schematics and images but appear to confirm those design changes. However, what remains unclear is whether the new ‌AirPods‌ will feature silicone ear-tips, like the ‌AirPods Pro‌, or feature the same in-ear design as the first and second-generation ‌AirPods‌.

Link

The New York Times Invites Beta Testers for its Audio App

Andrew Orr ·

The New York Times is building an app for audio journalism and it’s inviting beta testers to try it out.

“New York Times Audio” will provide an accessible and authoritative way to understand the world, pulling from Times podcasts, Times articles, premier magazine publishers like New York magazine and Rolling Stone, new audio formats from The Times newsroom and more. The product will also feature the archive of “This American Life,” encompassing 25 years worth of episodes from the iconic show that pioneered a new form of audio narrative journalism.

Link

Apple ML Study Compares Supervised Versus Self-Supervised Learning

Andrew Orr ·

A research team at Apple published a study in October examining supervised and self-supervised algorithms. The title is “Do Self-Supervised and Supervised Methods Learn Similar Visual Representations?” From the abstract:

We find that the methods learn similar intermediate representations through dissimilar means, and that the representations diverge rapidly in the final few layers. We investigate this divergence, finding that it is caused by these layers strongly fitting to the distinct learning objectives. We also find that SimCLR’s objective implicitly fits the supervised objective in intermediate layers, but that the reverse is not true.