Machine learning may be most useful in tiny, embedded, offline processors

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. Read the rest

Garbage In, Garbage Out: machine learning has not repealed the iron law of computer science

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. Read the rest

Nanoscale 3D printer that runs 100x faster than current models

A German startup called Nanoscribe says it will ship a nanoscale 3D printer in the second quarter of 2013, and that its device will run 100 times faster than similar devices currently in the market:

The technology behind most 3-D microprinters is called two-photon polymerization. It involves focusing tiny, ultrashort pulses from a near-infrared laser on a light-sensitive material. The material polymerizes and solidifies at the focused spots. As the laser beam moves in three dimensions, it creates a 3-D object.

Today’s printers, including Nanoscribe’s present system, keep the laser beam fixed and move the light-sensitive material along three axes using mechanical stages, which slows down printing. To speed up the process, Nanoscribe’s new tool uses a tiny moving mirror to reflect the laser beam at different angles. Thiel says generating multiple light beams with a microlens array could make the process even faster.

The smallest features that can be created using the Nanoscribe printer measure about 30 nanometers, says Julia Greer, professor of materials science at the California Institute of Technology.

Micro 3-D Printer Creates Tiny Structures in Seconds [Prachi Patel/MIT Technology Review]

(via Beyond the Beyond) Read the rest