Amazon Mechanical Turk

Amazon Mechanical Turk (MTurk) is a crowdsourcing website with which businesses can hire remotely located "crowdworkers" to perform discrete on-demand tasks that computers are currently unable to do as economically. It is operated under Amazon Web Services, and is owned by Amazon.^[1] Employers, known as requesters, post jobs known as Human Intelligence Tasks (HITs), such as identifying specific content in an image or video, writing product descriptions, or answering survey questions. Workers, colloquially known as Turkers or crowdworkers, browse among existing jobs and complete them in exchange for a fee set by the requester. To place jobs, requesters use an open application programming interface (API), or the more limited MTurk Requester site.^[2]As of April 2019, requesters could register from 49 approved countries.^[3] It is named after the Mechanical Turk, a chess machine.^[4]

The service was conceived by Venky Harinarayan in a U.S. patent disclosure in 2001.^[5][6][7] Amazon coined the term artificial artificial intelligence for processes that outsource some parts of a computer program to humans, for those tasks carried out much faster by humans than computers. MTurk product director Peter Cohen identified Jeff Bezos as the person who first proposed the concept.^[8]

The name Mechanical Turk was inspired by an 18th-century chess-playing automaton of the same name, often simply nicknamed as "The Turk". Made by Hungarian author and engineer Wolfgang von Kempelen, the machine became an international spectacle, touring Europe, and beating both Napoleon Bonaparte and Benjamin Franklin. It was later revealed that this "machine" was not an automaton, but rather controlled by a human chess master hidden in the cabinet beneath the board, puppeting the movements of a humanoid dummy. Analogously, the Mechanical Turk online service uses remote human labor hidden behind a computer interface to help employers perform tasks that are not currently possible using a true machine.

MTurk se lanzó en 2005. ^{[ 9 ]} En marzo de 2007, se informó que había más de 100 000 trabajadores en más de 100 países. ^{[ 10 ]} Esta cifra aumentó a más de 500 000 trabajadores registrados de más de 190 países en enero de 2011. ^{[ 11 ]} Ese año, Techlist publicó un mapa interactivo que señalaba las ubicaciones de 50 000 de sus trabajadores de MTurk en todo el mundo. ^{[ 12 ]} Para 2018, una investigación demostró que, si bien había más de 100 000 trabajadores disponibles en la plataforma en cualquier momento, solo alrededor de 2000 estaban trabajando activamente. ^{[ 13 ]}

Un usuario de Mechanical Turk puede ser un "Trabajador" (contratista) o un "Solicitante" (empleador). Los trabajadores tienen acceso a un panel que muestra tres secciones: ganancias totales, estado de la tarea y totales de tareas. Los trabajadores establecen sus propios horarios y no tienen la obligación de aceptar ninguna tarea en particular.

Amazon clasifica a los trabajadores como contratistas en lugar de empleados y no les paga impuestos sobre la nómina. Esta clasificación les permite evitar aspectos como el salario mínimo , las horas extras y la compensación laboral , una práctica común entre las plataformas de la economía colaborativa. En Estados Unidos, donde se concentra la gran mayoría de los trabajadores de MTurk, estos están obligados legalmente a declarar sus ingresos como autónomos . La diferente legalidad de esta práctica en países con leyes laborales más estrictas genera incertidumbre sobre la accesibilidad internacional del programa.

En 2013, el salario promedio por las múltiples microtareas asignadas, si se realizaban rápidamente, era de aproximadamente un dólar por hora, con cada tarea promediando unos pocos centavos. ^{[ 14 ]} Sin embargo, calcular las ganancias promedio por hora de las personas en un sitio de microtareas es extremadamente difícil y varias fuentes de datos muestran ganancias promedio por hora en el rango de $5 a $9 por hora ^{[ 15 ] [ 16 ] [ 17 ] [ 18 ]} entre un número sustancial de trabajadores, mientras que los trabajadores más experimentados, activos y competentes pueden ganar más de $20 por hora. ^{[ 19 ]}

Los trabajadores pueden tener una dirección postal en cualquier parte del mundo. El pago por completar las tareas se puede canjear en Amazon.com mediante un certificado de regalo (los certificados de regalo son la única opción de pago disponible para los trabajadores internacionales, excepto en India) o se puede transferir a una cuenta bancaria estadounidense del trabajador.

Los solicitantes pueden pedir a los trabajadores que cumplan con ciertos requisitos antes de realizar una tarea, y pueden establecer una prueba para verificar dicha cualificación. También pueden aceptar o rechazar el resultado enviado por el trabajador, lo cual afecta a su reputación. ( A partir de abril de 2019)Los solicitantes pagaban a Amazon una comisión mínima del 20% sobre el precio de los trabajos completados con éxito, con importes mayores para servicios adicionales . ^{[ 10 ]} Los solicitantes pueden usar la API de Amazon Mechanical Turk para integrar programáticamente los resultados del trabajo directamente en sus procesos y sistemas comerciales. Cuando los empleadores configuran un trabajo, deben especificar

• ¿Cuánto pagan por cada HIT completado?
• cuántos trabajadores quieren que trabajen en cada HIT,
• el tiempo máximo que un trabajador tiene para trabajar en una sola tarea,
• cuánto tiempo tienen los trabajadores para completar el trabajo,

así como los detalles específicos sobre el trabajo que desean que se realice.

The validity of research conducted with the Mechanical Turk worker pool has long been debated among experts.^[51] This is largely because questions of validity^[52][53] are complex: they involve not only questions of whether the research methods were appropriate and whether the study was well-executed, but also questions about the goal of the project, how the researchers used MTurk, who was sampled, and what conclusions were drawn.

Most experts agree that MTurk is better suited for some types of research than others. MTurk appears well-suited for questions that seek to understand whether two or more things are related to each other (called correlational research; e.g., are happy people more healthy?) and questions that attempt to show one thing causes another thing (experimental research; e.g., being happy makes people more healthy). Fortunately, these categories capture most of the research conducted by behavioral scientists, and most correlational and experimental findings found in nationally representative samples replicate on MTurk.^[54]

The type of research that is not well-suited for MTurk is often called "descriptive research." Descriptive research seeks to describe how or what people think, feel, or do; one example is public opinion polling. MTurk is not well-suited to such research because it does not select a representative sample of the general population. Instead, MTurk is a nonprobability, convenience sample. Descriptive research is best conducted with a probability-based, representative sample of the population researchers want to understand. When compared to the general population, people on MTurk are younger, more highly educated, more liberal, and less religious.^[55][22][36]

Mechanical Turk has been criticized by journalists and activists for its interactions with and use of labor. Computer scientist Jaron Lanier noted how the design of Mechanical Turk "allows you to think of the people as software components" in a way that conjures "a sense of magic, as if you can just pluck results out of the cloud at an incredibly low cost".^[56] A similar point is made in the book Ghostwork by Mary L. Gray and Siddharth Suri.^[57]

Critics of MTurk argue that workers are forced onto the site by precarious economic conditions and then exploited by requesters with low wages and a lack of power when disputes occur. Journalist Alana Semuels's article "The Internet Is Enabling a New Kind of Poorly Paid Hell" in The Atlantic is typical of such criticisms of MTurk.^[58]

Some academic papers have obtained findings that support or serve as the basis for such common criticisms,^[59][60] but others contradict them.^[61][62] A recent academic commentary argued that study participants on sites like MTurk should be clearly warned about the circumstances in which they might later be denied payment as a matter of ethics,^[63] even though such statements may not reduce the rate of careless responding.^[64]

A paper published by a team at CloudResearch^[18] shows that only about 7% of people on MTurk view completing HITs as something akin to a full-time job. Most people report that MTurk is a way to earn money during their leisure time or as a side gig. In 2019, the typical worker spent five to eight hours per week and earned around $7 per hour. The sampled workers did not report rampant mistreatment at the hands of requesters; they reported trusting requesters more than employers outside of MTurk. Similar findings were presented in a review of MTurk by the Fair Crowd Work organization, a collective of crowd workers and unions.^[65]

Mechanical Turk es comparable en algunos aspectos al servicio Google Answers, ahora descontinuado . Sin embargo, Mechanical Turk es una plataforma más general que puede ayudar a distribuir cualquier tipo de tarea laboral en todo el mundo. El Intérprete Humano Colaborativo (CHI, por sus siglas en inglés), de Philipp Lenssen, también propuso utilizar inteligencia humana distribuida para ayudar a los programas informáticos a realizar tareas que las computadoras no pueden hacer bien. MTurk podría utilizarse como motor de ejecución para el CHI.

En 2014, el gigante ruso de búsquedas Yandex lanzó un sistema similar llamado Toloka , similar al Turco Mecánico. ^{[ 82 ]}

• CAPTCHA , que desafía y verifica el trabajo humano en una tarea en línea sencilla
• Ciencia ciudadana
• Microtrabajo
• Amazon Go

• Business Week article on Mechanical Turk by Rob Hof, November 4, 2005.
• Wired Magazine story about "Crowdsourcing," June 2006.
• Salon.com article on Mechanical Turk by Katharine Mieszkowski, July 24, 2006.
• New York Times article on Mechanical Turk by Jason Pontin, March 25, 2007.
• Technology Review article on Mechanical Turk, "How Mechanical Turk is Broken," by Christopher Mims, January 3, 2010.
• J. Nathan Matias (June 8, 2015), "Tragedy of the Digital Commons", The Atlantic (discusses labor relations)

• Official website
• Requester Best Practices Guide, Updated February 2015.
• Matt Lease (ed.). "Amazon Mechanical Turk". Crowdsourcing News, Events, and Resources. US – via University of Texas at Austin School of Information.