noncomputablenotes - Tumblr blog

[noncomputablenotes]

Cooperation emerges when groups are small and memories are long, study finds

https://phys.org/news/2016-06-cooperation-emerges-groups-small-memories.html

[noncomputablenotes]

Habitus

https://twitter.com/Aelkus/status/855954357338222592

Trolls lack the Habitus, or social/cultural “finger tips feeling” to refine sentiments into acceptability for mainstream society, and so absorb culture and turn it into what may be a more honest or “implicative” form as they spew it back out.

Different habituses are stabilizing as they can keep problematic subcultures from gaining influence, but sometimes the mainstream institutions break this divide and normalize problematic subcultural norms (i.e. MSM embracing Tumblrisms).

[noncomputablenotes]

In the Swarm

https://mitpress.mit.edu/books/swarm

How “swarms” are distinct from “crowds” and “masses” and other notions usual in looking at society and social change. Swarms differ in that they are transient;  they do not form any enduring whole or collective of interests, no “we”, but merely a collection of individuals that temporarily direct themselves towards an object of outrage.

[noncomputablenotes]

Liberty and Consequentialism

Liberty and Consequentialism

http://rationallyspeaking.blogspot.com/2010/11/on-utilitarianism-and-consequentialism.html

http://julessearchforvirtue.blogspot.com/2010/02/liberty-and-consequentialism.html

[noncomputablenotes]

Steve Ballmer Serves Up a Fascinating Data Trove

https://www.nytimes.com/2017/04/17/business/dealbook/steve-ballmer-serves-up-a-fascinating-data-trove.html?_r=0

Steve Ballmer has been undergoing a hushed down venture since retiring from his position as CEO of Microsoft (and purchasing a basketball team). He has hired economists and other such experts to gather as much data directly from the government as possible to understand exactly what money is spent on, how resources are allocated, and where taxes go. People will be able to use the data to help themselves form their own political opinions more effectively, it can be used to generate customized reports, and (I imagine) perhaps applications can be developed over the data.

It is called USAFacts, at http://usafacts.org.

[noncomputablenotes]

Twelve types of Artificial Intelligence (AI) problems

https://opendatascience.com/blog/twelve-types-of-artificial-intelligence-ai-problems/

General overview of 12 different (but sometimes intersecting) types of problems that can be solved with AI (namely, deep learning).

[noncomputablenotes]

Why did Google dump Boston Dynamics?

http://globalguerrillas.typepad.com/globalguerrillas/2016/03/why-did-google-dump-boston-dynamics.html There was a culture clash after Google acquired Boston Dynamics, with the BD team disinclined from the collaboration with Google's American and Japanese engineers. According to a deep learning specialist, a friend of the author, one big trouble was that BD did not want to implement deep learning—they'd been using specialized models, and this was the working of their engineers. Google may have proposed this, to go from efficiently-engineered robots to flexible deep-learning optimized ones, BD may have given too long a timeframe for training and redevelopment, and finally Google may have decided that the way to go was not with Boston Dynamics and that it'd be best to sell it (perhaps while keeping the important hardware-related patents).

[noncomputablenotes]

The Importance of Location

https://mapr.com/blog/importance-location-real-estate-weather-and-machine-learning/#.U13dOvldUt4

Explains the importance of boundary areas between classes of data in feature spaces by making an analogy to regional weather prediction and real estate.

[noncomputablenotes]

Unsettling Web Ads

http://www.remarkably.com/secret-story-weird-images-changing-face-adverts/

http://swiped.co/file/eat-this-never-diet-again-wierd-image-banner-ads/

Those unsettling pictures that appear on “YOU WON’T BELIEVE...”, “ONE QUICK TRICK...”, “DO THIS, YOU”LL NEVER HAVE TO...” ads on websites (often online magazines) may well be intended to do two things:

1) Filter out people skeptical enough to avoid, filter in people gullible enough to believe. If they’re so gullible, they’ll probably be gullible to fall for future scams as well.

2) They present weird, obscure objects and scenes that incline people to click on them out of curiosity. The companies peddling them may have zero to negative reputation and so may actually gain from doing this and drawing people in more than they’d lose from the reputation that comes with pushing these images (unlike, say, large enterprises with established positive or growing reputations).

[noncomputablenotes]

Facebook’s VR

https://www.bloomberg.com/features/2016-facebook-virtual-reality/

Facebook, with Mark Zuckerberg and Oculus founders like Palmer Luckey at the helm, is pursuing Virtual Reality as the defining technology of the future. The key, they think, to getting billions of users is to start through mobile VR. They need long term funding and lots of science on perception and sensing, and are cruising forward.

[noncomputablenotes]

What’s the difference between an “urban traveler” and a homeless kid? Not much.

https://www.denverite.com/what-is-an-urban-traveler-7487/

Urban travelers are often young people with troubles, from the home (or foster home), drug abuse, mental illness, and the like. Many show symptoms of PTSD, may be relatively polite compared to “native” homeless, and have a high propensity to smoke marijuana; though, they may still be rude due to high rates of PTSD and still have a problematic propensity for violence. They’re often “runaways”. The city (Denver, but maybe others too) calls them “urban travelers” but they are more likely to call themselves just “travelers” or even “gypsies”.

[noncomputablenotes]

Who says Programming isn’t creative?

https://salfreudenberg.wordpress.com/2013/07/09/who-says-programming-isnt-creative/

Study on how expert programmers use complex, multidimensional visualizations to explore programming problems, figure errors, move back and forth in process, etc.

[noncomputablenotes]

Watchlist: data brokers

http://digitalrightswatch.org.au/2016/10/18/watchlist-data-brokers/

Large companies (”corporations”) from social media websites to online retailers to physical retailers with rewards/loyalty cards/programs to credit card companies sell and give out access to bits of data, often anonymized, which companies called data brokers buy, aggregate, analyze, and connect. This data is then sold to organizations like stores and political parties to help them target efforts, like persuasion efforts.

[noncomputablenotes]

Second Life

https://motherboard.vice.com/en_us/article/why-is-second-life-still-a-thing-gaming-virtual-reality

History of the biggest open-world customizable 3D world-building social network and its erotic hubs.

[noncomputablenotes]

Jews and African Americans

http://scholarsarchive.byu.edu/cgi/viewcontent.cgi?article=4573&context=etd

http://www.kevinmacdonald.net/jews&blacks.pdf

[noncomputablenotes]

Movies with that settluncanny feeling

Common features: Grainy in sound and visual, (sub)urban, teen flick, troubled degenerate teens

American Beauty

The Virgin Suicides

The Breakfast Club

Gimme Shelter

Lords of Dogtown

Kids (haven’t watched yet)

Donnie Darko (maybe)

Gummo (potentially, and other movies by Harmony Korine seem to have a similar tone)

[noncomputablenotes]

Notes From The Asilomar Conference On Beneficial AI

http://slatestarcodex.com/2017/02/06/notes-from-the-asilomar-conference-on-beneficial-ai/

SSC went to an AI conference.

Looking back at human strategic game culture and seeing where turning points which could’ve found us with the strategies derived by AI were missed, for use in optimizing our own strategic culture and possibly beating AI’s in the future.

Treating reinforcement learners ethically.

Deriving human values and moral guidelines from watching humans.

Problems with nontransparency of AI’s.