Pete Warden (previously) is one of my favorite commentators on machine learning and computer science; yesterday he gave a keynote at the IEEE Custom Integrated Circuits Conference, on the ways that hardware specialization could improve machine learning: his main point is that though there's a wealth of hardware specialized for creating models, we need more hardware optimized for running models.

Pete Warden (previously) writes persuasively that machine learning companies could make a ton of money by turning to data-compression: for example, ML systems could convert your speech to text, then back into speech using a high-fidelity facsimile of your voice at the other end, saving enormous amounts of bandwidth in between.

The tiny embedded processors in smart gadgets — including much of the Internet of Shit — are able to do a lot of sensing without exhausting their batteries, because sensing is cheap in terms of power consumption.

Pete Warden writes convincingly about computer scientists' focus on improving machine learning algorithms, to the exclusion of improving the training data that the algorithms interpret, and how that focus has slowed the progress of machine learning.

Machine learning is often characterized as much an "art" as a "science" and in at least one regard, that's true: its practitioners are prone to working under loosely controlled conditions, using training data that is being continuously tweaked with no versioning; modifying parameters during runs (because it takes too long to wait for the whole run before making changes); squashing bugs mid-run; these and other common practices mean that researchers often can't replicate their own results — and virtually no one else can, either.

Pete Warden reports in from the ARM Research Summit, where James Myers presented on "energy harvesting" by microscopic computers — that is, using glints of sunlight and the jostling of motion from bumping into things or riding on our bodies to provide power for computation.

Pete Warden writes on O'Reilly Radar about the problems of anonymizing datasets. AOL, Netflix and others have been burned by releasing datasets that they thought had been stripped of identifiable elements, only to discover that de-anonymizing some or all of the data was easier than they thought. — Read the rest

Surely you've been following the iPhone-tracks-your-location-data story in recent weeks. As of today, the Wall Street Journal is the latest large outlet to cover the story of security experts Alasdair Allan and Pete Warden's study of what location data the device stores about where users go, on what dates. — Read the rest

Security researchers presenting at the Where 2.0 conference have revealed a hidden, secret iOS file that keeps a record of everywhere you've been. The record is synched to your PC and subsequently resynched to your other mobile devices. The file is not transmitted to Apple, but constitutes a substantial privacy breach if your PC or mobile device are lost or seized. — Read the rest