What Was I Just Computing Again

Forget about it: A material that mimics the brain
Oct 2017, phys.org

Lattice breathing, electronic forgetting, and proton doping, oh my.

It's hard to find a material that forgets, but they've found one (U.S. Department of Energy's (DOE) Argonne National Laboratory, in collaboration with others), and now they're trying to make a better computer by making it forget, because we forget, and we are the ultimate computing machine, despite what you might think these days.

Post Script - On Forgetting Again

The internet is rotting - Thousands of sites go offline each year
July 2019, phys.org

How do we remove biases in AI systems? Start by teaching them selective amnesia
Mar 2020, phys.org

Jaiswal and co-author Daniel Moyer, Ph.D., developed the adversarial forgetting approach, which teaches deep learning models to disregard specific, unwanted data factors so that the results they produce are unbiased and more accurate.
...
Deep learning algorithms are great at learning things, but it's more difficult to make sure that the algorithms don't learn certain things. Developing algorithms is a very data-driven process, and data tends to contain biases.

On Crowdsourcing

wondergrounder

Ford challenges developers to create efficiency app
Alisa Priddle, March 29, 2013 

[...seems to be more about Ford using its consumers to do 'real-world' research and development, using Apple and Google products which the consumers themselves also pay for, of course...I call that freesearch]

Ford is offering up to $50,000 to developers who deliver hardware or software that helps drivers understand the effect of the elements and driving habits on their fuel economy.

"We need to help customers understand the concept of personal fuel economy, based on their own individualized experiences, and give them tools to see, learn and act upon all the information available to know what to expect, how to improve, and even offer guidance in their shopping process," Farley said.

The announcement follows consumer complaints, class action lawsuits and a U.S. Environmental Protection Agency investigation into the gap between mileage posted on new vehicle stickers and what drivers claim they are getting.

BONUS [see "real-world"]
Ford has said it is working with the EPA and discussing whether the government tests need to be revised to better reflect real-world conditions.

More Algos

copyright bots fail

The Algorithmic Copyright Cops: Streaming Video’s Robotic Overlords

BY GEETA DAYAL 09.06.12

"As live streaming video surges in popularity, so are copyright “bots” — automated systems that match content against a database of reference files of copyrighted material. These systems can block streaming video in real time, while it is still being broadcast, leading to potentially worrying implications for freedom of speech."

The volume of content is overwhelming, encouraging the use of automated filters, working via video/audio-monitoring algorithms.

...article goes on to mention various embarissing incidents where public feeds (by any definition) were automatically suspended by autonomous bots, which are apparently not 'smart' enough to see what is quite obvious to us humans.

"Given all that, it’s likely that this collision between algorithmic defense of copyright versus spontaneous speech isn’t going to be resolved soon."

http://www.wired.com/threatlevel/2012/09/streaming-videos-robotic-overlords-algorithmic-copyright-cops/

"Intellectual Property"

Electronic Frontier Foundation

POST SCRIPT:

YouTube Upgrades Its Automated Copyright Enforcement System | Electronic Frontier Foundation


M.I.T. Computer Program Reveals Invisible Motion in Video
[video link]
Eulerian Video Magnification for Revealing Subtle Changes in the World
[via]


So It Begins: Darpa Sets Out to Make Computers That Can Teach Themselves
BY ROBERT BECKHUSEN, 03-2013
“Probabilistic Programming for Advanced Machine Learning”
When Darpa talks about artificial intelligence, it’s not talking about modeling computers after the human brain. That path fell out of favor among computer scientists years ago as a means of creating artificial intelligence; we’d have to understand our own brains first before building a working artificial version of one. But the agency thinks we can build machines that learn and evolve, using algorithms — “probabilistic programming” — to parse through vast amounts of data and select the best of it. After that, the machine learns to repeat the process and do it better.