Every time Apple said 'machine learning', we had a drink andsgd oh*][

Yasss Steve i mean Tim... sorry... tell ush moar ab, ab, aboat aaye eye


WWDC While touting forthcoming operating system features at its annual developer conference on Monday, Apple made sure to mention machine learning and related AI-oriented terminology over and over.

Kevin Lynch, technology veep, talked about Siri, Apple's personal assistant software, becoming more proactive and more aware of watchOS activity through machine learning.

Craig Federighi, senior veep of software engineering, highlighted Safari's use of machine learning for intelligent blocking of browser tracking. He also talked about advanced convolutional neural networks improving facial recognition in Photos and making Siri smarter.

Federighi cited the utility of Apple's new Metal 2 graphics API for machine learning. And he said deep learning had been used to make Siri's voice sound more natural.

John Ternus, veep of hardware engineering, called out the compute power of Apple's forthcoming iMac Pro by noting how its Radeon Vega GPU would be helpful for machine learning.

Like alumni of prestigious schools who work their alma mater into every conversation, Apple execs made certain everyone was aware that their machines have learned.

A year ago, developer Marco Arment fretted about Apple's inaction on AI in a blog post. If the bets being made on AI by Amazon, Facebook, and Google pay off, Arment mused, "Apple will find itself in a similar position as BlackBerry did almost a decade ago: What they’re able to do, despite being very good at it, won’t be enough anymore, and they won’t be able to catch up."

AI isn't new to Apple, which has been nurturing Siri since 2011, but it didn't become a priority for the company until last October, when it hired Carnegie Mellon University professor Ruslan Salakhutdinov as its first director of AI research.

Apple in the past had trouble attracting accomplished AI researchers because its culture of secrecy rubbed academics the wrong way. It's not coincidental that Apple published its first AI research paper last December. And the following month, Apple joined the Partnership on AI, an industry consortium established last year to promote best practices.

Now, not only is Apple underscoring its use of AI-ish tech, but the company is bringing the technology it has been using for its Siri and Camera applications to developers. Its forthcoming platform updates – watchOS 4, tvOS 11, macOS 10.13, and iOS 11 – provide access to Core ML, a machine learning framework that includes a Vision API and a Natural Language API.

It's as if Apple is trying to tell us something. ®

Similar topics


Other stories you might like

  • Apple strays from the path of locking down parts with its Series 7 Watch

    Component swaps still a thing... for now

    Apple's seventh-gen Watch has managed to maintain its iFixit repairability rating on a par with the last model – unlike its smartphone sibling.

    The iFixit team found the slightly larger display of the latest Apple Watch a boon for removal via heat and a suction handle. Where the previous generation required a pair of flex folds in its display, the new version turned out to be simpler, with just the one flex.

    Things are also slightly different within the watch itself. Apple's diagnostic port has gone and the battery is larger. That equates to a slight increase in power (1.094Wh from 1.024Wh between 40mm S6 and 41mm S7) which, when paired with the slightly hungrier display, means battery life is pretty much unchanged.

    Continue reading
  • Better late than never: Microsoft rolls out a public preview of E2EE in Teams calls

    Only for one-to-one voice and video, mind

    Microsoft has finally kicked off the rollout of end-to-end-encryption (E2EE) in its Teams collaboration platform with a public preview of E2EE for one-to-one calls.

    It has been a while coming. The company made the promise of E2EE for some one-to-one Teams calls at its virtual Ignite shindig in March this year (https://www.theregister.com/2021/03/03/microsoft_ups_security/) and as 2021 nears its end appears to have delivered, in preview form at least.

    The company's rival in the conference calling space, Zoom, added E2EE for all a year ago, making Microsoft rather late to the privacy party. COO at Matrix-based communications and collaboration app Element, Amandine Le Pape, told The Register that the preview, although welcome, was "long overdue."

    Continue reading
  • Recycled Cobalt Strike key pairs show many crooks are using same cloned installation

    Researcher spots RSA tell-tale lurking in plain sight on VirusTotal

    Around 1,500 Cobalt Strike beacons uploaded to VirusTotal were reusing the same RSA keys from a cracked version of the software, according to a security researcher who pored through the malware repository.

    The discovery could make blue teams' lives easier by giving them a clue about whether or not Cobalt Strike traffic across their networks is a real threat or an action by an authorised red team carrying out a penetration test.

    Didier Stevens, the researcher with Belgian infosec firm NVISO who discovered that private Cobalt Strike keys are being widely reused by criminals, told The Register: "While fingerprinting Cobalt Strike servers on the internet, we noticed that some public keys appeared often. The fact that there is a reuse of public keys means that there is a reuse of private keys too: a public key and a private key are linked to each other."

    Continue reading

Biting the hand that feeds IT © 1998–2021