Towards a better understanding of Social Machines

Towards a better understanding of Social Machines

Alvaro Graves

Outline

Part 1: Cognitive processes involve (sometimes) the use of the Web

Part 2: The Web can (sometimes) be enhanced by the human mind

Part 3: Our current initiatives to study this interaction

<part1><!-- The Web in Cognition -->

Cognition is not onlyin the mind

• The Extended Mind (Clark & Chalmers, 1998)

– “... beliefs can be constituted partly by features of the environment”

– Otto's notebook

• Organism-Centered Cognition (Clark, 2007)

– Cognitive process extended to the environment

– However, centered in the organism

Cognition and the Web

• Cognitive Extension and the Web. (Smart et al., 2009)

Otto's Web-based notebook?

“Wikipedia is down, I'm ignorant again”A. Cadiz, 2010

Conditions for Cognition

If Web is part of Cognition, several conditions must be fulfilled:

Availability (reliable and invoked frequently) Trust (information as trustworthy as bio-memory) Accessibility (can obtain information efficiently) Conscious endorsement (information endorsed

in the past)

Availability criterion

• More and more people take action after consulting the Web

• Increased ubiquity of the Web– Laptops– Cellphones– Game consoles

– TV– Refrigerators

Trust criterion

• Google Maps?– Grocery shops in Troy, NY.– GMaps shows me how to get there

• Medical information from PubMed?

Accessibility criterion

• “Wristwatch example” (Clark, 2003)

– Do you know what time is it?

• Information doesn't need to be consciously known

• However is important to being able to (efficiently) access it.

• Access Wikipedia?

Conscious endorsement

• A weak criterion (Clark, 2008)

–What about implicit knowledge?

• User-provided content– Revisiting articles– My Weblog

Conclusions part 1

• Can be part of cognitive processes– Not all the Web– Not all the time

• However it seems this is becoming more and more part of it.

</part1>

<part2><!-- Cognition in the Web -->

Motivation

Lots of human computational power

9 billion human-hours of solitaire were played in 2003

– Empire state building, 7 million human-hours (6.8 hours of Solitare)

– Panama canal, 20 million human-hours (less than a day of Solitare)

Social Machines

• Social Machines are mechanisms where:

– Humans do the creative work

– Machines do the administrative work

What are we good at?

Humans Computers

Discover patterns

Good Bad

Creative thinking

Good Bad

Information Management

Bad Good

Data communication

Bad Good

Types of Social Machines

• There is a range of mechanisms available, but we can classify them in two groups (Haythornthwaite, 2009)

– Heavyweight

– Lightweight

Heavyweight

Smaller audience, long-term commitment, democratic/meritocratic

•Examples:

–Wikipedia

– F/OSS projects

–W3C Working groups

– Fansub groups

Lightweight

Bigger audience, short-term commitment, non-democratic

Examples:– Fold.it (http://fold.it)– ReCaptcha (http://recaptcha.net/)– GalaxyZoo

(http://www.galaxyzoo.org/)

Conclusions part 2

Different types of Social Machines Heavyweight Lightweight

Human cognition is key in these Social Machines

</part2>

<part3><!-- How to study SM -->

How to study this phenomenon?

Limitations Access to information (logs, database) Privacy concerns

Solution Create framework for Social Machines Incentives are important

Motivation 1:Public Safety

Interest for individuals, policy makers, law-enforcers

Provide “official” information (TroyPD, RPI public safety)

Also allow users to include relevant information

Enable establishing trust relations between users and belief on data

PublicSafetyMap.org

Next steps

Annotation on events

Ability to put ANYTHING with geolocation

Connect with well-known social networks (Facebook, MySpace, etc...)

Motivation 2: Geoannotation

User can mark areas (polygons) in a map (Sage bldg., my office, home, etc...)

Allow users to operate and search over these polygons

“Give me all chinese restaurant that delivers to my place”

Conclusions part 3

Don't compete with Google

Opening data so others can use it for their own applications

But most important: A Framework where we can study Social Machines

</part3>

<conclusions>

Conclusions

Cognition is (sometimes) related to the Web

Web2.0 has made available incipient Social Machines

What can we do with Web3.0 technologies?

How to improve trust, collaboration, privacy, efficiency in this and new SM's?

</conclusions><questions/>

<references/>

Clark and Chalmers. The extended mind. Analysis (1998) vol. 58 (1) pp. 7-19

Clark. Curing cognitive hiccups: A defense of the extended mind. (2007)

Smart et al. Cognitive Extension and the Web. (2009) Clark. Natural-born cyborgs: Minds, technologies, and the

future of human intelligence. (2003) Clark. Supersizing the mind: embodiment, action, and cognitive

extension, 2008 Hendler and Berners-Lee. From the Semantic Web to social

machines: A research challenge for AI on the World Wide Web. Artificial Intelligence (2010) vol. 174 (2) pp. 156-161