Human-based computation: Difference between revisions

Revision as of 19:24, 4 January 2023

Revision as of 22:09, 29 January 2023

Line 1:

{{lead too short|date=October 2022}}

{{Use mdy dates|date=September 2017}}

'''Human-based computation''' ('''HBC'''), '''human-assisted computation''',<ref>{{cite web |url=https://commonsensereasoning.org/2007/papers/shahaf-and-amir.pdf |title=Towards a Theory of AI Completeness |author=Shahaf, Dafna; and Eyal Amir |date=28 March 2007 |access-date=12 May 2022}}</ref> '''ubiquitous human computing''' or '''distributed thinking''' (by analogy to [[distributed computing]]) is a [[computer science]] technique in which a machine performs its function by outsourcing certain steps to humans, usually as [[microwork]]. This approach uses differences in abilities and alternative costs between humans and computer agents to achieve symbiotic human–computer interaction. For computationally difficult tasks such as image recognition, human-based computation plays a central role in training [[Deep Learning]]-based [[Artificial intelligence|Artificial Intelligence]] systems. In this case, human-based computation has been referred to as '''human-aided artificial intelligence'''.<ref>{{Cite journal|last=Mühlhoff|first=Rainer|date=2019-11-06|title=Human-aided artificial intelligence: Or, how to run large computations in human brains? Toward a media sociology of machine learning |journal=New Media & Society |volume=22|issue=10|language=en|pages=1868–1884|doi=10.1177/1461444819885334|s2cid=209363848|issn=1461-4448|url=https://depositonce.tu-berlin.de/handle/11303/12510|doi-access=free}}</ref>

In traditional computation, a human employs a computer<ref>the term "computer" is used the modern usage of computer, not the one of [[human computer]]</ref> to solve a problem; a human provides a formalized problem description and an algorithm to a computer, and receives a solution to interpret.<ref>{{cite web|url=https://www.csee.umbc.edu/courses/471/papers/turing.pdf|author=Turing, Alan M.|title=Computer Machinery and Intelligence|date=1950|access-date=12 May 2022}}</ref> Human-based computation frequently reverses the roles; the computer asks a person or a large group of people to solve a problem,<ref>{{cite book|chapter-url=https://ieeexplore.ieee.org/document/1225961?section=abstract|chapter=Automatic concept evolution|author=Fogarty, Terence C.|title=The Second IEEE International Conference on Cognitive Informatics, 2003. Proceedings. |date=20 August 2003|page=89 |doi=10.1109/COGINF.2003.1225961 |isbn=0-7695-1986-5 |s2cid=30299981 |access-date=21 June 2021}}</ref> then collects, interprets, and integrates their solutions. This turns hybrid networks of humans and computers into "large scale distributed computing networks".<ref>{{Citation|last=von Ahn|first=Luis|title=Human Computation |url=https://www.youtube.com/watch?v=tx082gDwGcM |archive-url=https://ghostarchive.org/varchive/youtube/20211219/tx082gDwGcM |archive-date=2021-12-19 |url-status=live |volume=Google Tech Talk July 26, 2006 |access-date=2019-11-22}}{{cbignore}}. Cited after Mühlhoff, Rainer (2019). [https://journals.sagepub.com/doi/abs/10.1177/1461444819885334 "Human-aided artificial intelligence: Or, how to run large computations in human brains? Toward a media sociology of machine learning".] New Media & Society: 146144481988533. doi:10.1177/1461444819885334. ISSN 1461-4448.</ref><ref>{{cite web |url=http://fc05.ifca.ai/p26.pdf|title=Secure Distributed ''Human'' Computation |author=Gentry, Craig; Zulfikar Ramzan, and Stuart Stubblebine |access-date=12 May 2022}}</ref><ref>{{cite book |chapter-url=https://link.springer.com/chapter/10.1007/11507840_28 |title=Secure Distributed ''Human'' Computation |chapter=Secure Distributed Human Computation |series=Lecture Notes in Computer Science |year=2005 |doi=10.1007/11507840_28 |access-date=12 May 2022|last1=Gentry |first1=Craig |last2=Ramzan |first2=Zulfikar |last3=Stubblebine |first3=Stuart |volume=3570 |pages=328–332 |isbn=978-3-540-26656-3 }}</ref> where code is partially executed in human brains and on silicon based processors.

Line 11:

A concept of the automatic [[Turing test]] pioneered by [[Moni Naor]] (1996)<ref>{{cite web |url=http://www.wisdom.weizmann.ac.il/~naor/PAPERS/human_abs.html |author=Naor, Moni |title=Verification of a human in the loop or Identification via the Turing Test |access-date=12 May 2021}}</ref> is another precursor of human-based computation. In Naor's test, the machine can control the access of humans and computers to a service by challenging them with a [[natural language processing]] (NLP) or [[computer vision]] (CV) problem to identify humans among them. The set of problems is chosen in a way that they have no algorithmic solution that is both effective and efficient at the moment. If it existed, such an algorithm could be easily performed by a computer, thus defeating the test. In fact, Moni Naor was modest by calling this an automated Turing test. The [[imitation game]] described by [[Alan Turing]] (1950) didn't propose using CV problems. It was only proposing a specific NLP task, while the Naor test identifies and explores a large [[AI-complete|class]] of problems, not necessarily from the ___domain of NLP, that could be used for the same purpose in both automated and non-automated versions of the test.

Finally, [[Human-based genetic algorithm]] (HBGA)<ref>{{cite book |chapter-url=https://ieeexplore.ieee.org/document/972056 |title=Human-based genetic algorithm |chapter=Human based genetic algorithm |year=2001 |doi=10.1109/ICSMC.2001.972056 |access-date=12 May 2022|last1=Kosorukoff |first1=A. |volume=5 |pages=3464–3469 |isbn=0-7803-7087-2 |s2cid=13839604 }}</ref> encourages human participation in multiple different roles. Humans are not limited to the role of evaluator or some other predefined role, but can choose to perform a more diverse set of tasks. In particular, they can contribute their innovative solutions into the evolutionary process, make incremental changes to existing solutions, and perform intelligent recombination.<ref>{{cite web |url=http://gpbib.cs.ucl.ac.uk/gecco2005lbp/papers/56-hammond.pdf |author=Hammond, Michelle O.; and Terence C. Fogarty|title=Co-operative OuLiPian (Ouvroir de littérature potentielle) Generative Literature Using Human-Based Evolutionary Computing |access-date=12 May 2022}}</ref> In short, HBGA allows humans to participate in all operations of a typical [[genetic algorithm]]. As a result of this, HBGA can process solutions for which there are no computational innovation operators available, for example, natural languages. Thus, HBGA obviated the need for a fixed representational scheme that was a limiting factor of both standard and interactive EC.<ref>{{cite journal |url=https://ieeexplore.ieee.org/document/949485 |title=Interactive evolutionary computation: fusion of the capabilities of EC optimization and human evaluation, pp.&nbsp;1275-1296 |author=Takagi, Hideyuki |journal=Proceedings of the IEEE |date=September 2001 |volume=89 |issue=9 |pages=1275–1296 |doi=10.1109/5.949485 |hdl=2324/1670053 |access-date=12 May 2022|hdl-access=free }}</ref> These algorithms can also be viewed as novel forms of social organization coordinated by a computer, according to Alex Kosorukoff and David Goldberg.<ref>{{cite web |url=http://research.3form.com/alex/pub/gecco-2002-18.pdf |title=Evolutionary Computation as a Form of Organization, pp.&nbsp;965-972 |access-date=12 May 2022 |archive-url=https://web.archive.org/web/20110707063732/http://research.3form.com/alex/pub/gecco-2002-18.pdf |archive-date=7 July 2011 |url-status=dead}}</ref>

==Classes of human-based computation==

Line 95:

Human-based computation has been criticized as exploitative and deceptive with the potential to undermine collective action.<ref>{{cite web |url=https://cyber.harvard.edu/events/2010/02/zittrain |title=Minds for Sale |author=Zittrain, Jonathan |date=July 20, 2019 |access-date=12 May 2022}}</ref><ref>{{cite book |url=https://dl.acm.org/doi/10.1145/2132176.2132186 |title=Exploring the character of participation in social media: the case of Google Image Labeler |author=Jafarinaimi, Nassim |date=February 7, 2012 |pages=72–79 |doi=10.1145/2132176.2132186 |isbn=9781450307826 |s2cid=7094199 |access-date=12 May 2022}}</ref>

In [[social philosophy]] it has been argued that human-based computation is an implicit form of online labour.<ref>{{cite journal |url=https://journals.sagepub.com/doi/full/10.1177/1461444819885334|title=Human-aided artificial intelligence: Or, how to run large computations in human brains? Toward a media sociology of machine learning |author=Mühlhoff, Rainer |journal=New Media & Society |year=2020 |volume=22 |issue=10 |pages=1868–1884 |doi=10.1177/1461444819885334 |s2cid=209363848 |access-date=12 May 2022|doi-access=free }}</ref> The philosopher Rainer Mühlhoff distinguishes five different types of "machinic capture" of human microwork in "hybrid human-computer networks": (1) gamification, (2) "trapping and tracking" (e.g. CAPTCHAs or click-tracking in Google search), (3) social exploitation (e.g. tagging faces on Facebook), (4) information mining and (5) click-work (such as on [[Amazon Mechanical Turk]]).<ref>{{Cite journal|last=Mühlhoff|first=Rainer|date=2019-11-06|title=Human-aided artificial intelligence: Or, how to run large computations in human brains? Toward a media sociology of machine learning|journal=New Media & Society|volume=22|issue=10|language=en|pages=1868–1884|doi=10.1177/1461444819885334|s2cid=209363848|issn=1461-4448|url=https://depositonce.tu-berlin.de/handle/11303/12510|doi-access=free}}</ref><ref>{{cite web|url=https://philpapers.org/archive/MHLHAI-2.pdf |title=Human-aided artificial intelligence: Or, how to run large computations in human brains? Toward a media sociology of machine learning |author=Mühlhoff, Rainer |access-date=12 May 2022}}</ref> Mühlhoff argues that human-based computation often feeds into [[Deep learning|Deep Learning]]-based [[Artificial intelligence|Artificial Intelligence]] systems, a phenomenon he analyzes as "human-aided artificial intelligence".

==See also==