by Karen Hao

The artificial-intelligence industry runs on the invisible labor of humans working in isolated and often terrible conditions—and the model is spreading to more and more businesses.

On Wednesday, the Guardian published an article about the realities of producing Google Assistant. Behind the “magic” of its ability to interpret 26 languages is a huge team of linguists, working as subcontractors, who must tediously label the training data for it to work. They earn low wages and are routinely forced to work unpaid overtime. Their concerns over working conditions have been repeatedly dismissed.

It’s just one story among dozens that have begun to peel back the curtain on how the artificial-intelligence industry operates. Human workers don’t just label the data that makes AI work. Sometimes humans workers are the artificial intelligence. Behind Facebook’s content-moderating AI are thousands of content moderators; behind Amazon Alexa is a global team of transcribers; and behind Google Duplex are sometimes very human callers mimicking the AI that mimics humans. Artificial intelligence doesn’t run on magic pixie dust. It runs on invisible laborers who train algorithms relentlessly until they’ve automated their own jobs away.

In their new book Ghost Work: How to Stop Silicon Valley from Building a New Global Underclass, anthropologist Mary Gray and computer scientist Siddharth Suri argue that you and I could be next.

I caught up with Gray this week to discuss why people turn to ghost work, how their invisibility leaves them more vulnerable to terrible working conditions, and how we can make this new form of work more sustainable.

The following has been edited for length and clarity.

MIT Technology Review: How do you define ghost work?

10 JUNE 2019 - 20:03  MUDIWA GAVAZA

The report suggests that companies need to implement programmes to reskill their human capital for the changing business environment

The introduction and use of artificial intelligence (AI) in SA may be hampered by concerns around job losses but experts argue new prospects could be created by the same technology, says a new report by EY Africa.

“Currently, 54% of companies rate impact on personnel as the top business risk in implementing AI,” said the AI Maturity Study, which was commissioned by Microsoft and conducted by EY.

AI has to do with machines demonstrating intelligence normally displayed by humans such as learning and problem solving.

With SA’s official unemployment above 27%, fear of potential job losses has dominated discussion around AI in business.

“In a region where unemployment rates are high, the job market is highly unionised and true understanding of AI is low, employees have an inherent reservation towards scaled adoption of AI and significant amounts of education and change management will be required,” said the report, released on Monday.

To fight this “automation anxiety”, the report suggests that companies need to implement programmes to reskill their human capital for the changing business environment. 

Microsoft SA’s managing director Lillian Barnard said: “I think we can take a cue from past major technological innovations like the invention of the telephone and the automobile.” There was a shift then and we should expect a shift now.

“It’s hard to predict what the future will bring. What we need is to focus our approach on providing the skills and training to ensure that we balance the equation. As much as we may perhaps lose jobs on one side, we can create new jobs on the other,” said Barnard. 

07.06.2019

European Commission awards Catalan capital one of three new supercomputers in 2021

The Barcelona Supercomputing Center (BSC) has been chosen to host and operate one of the most powerful supercomputers in Europe from 2021, the European Commission has announced.

The BSC, with the support of five EU member states as well as the Catalan government, won the commission alongside teams from Finland and Italy as part of a pioneering continent-wide plan to compete with the USA, China and Japan as major international players in supercomputing.

The EU has pledged a total of 840 million euros – including its largest-ever research grant to the Spanish state of 200 million euros – for the project, which will contribute to building a supercomputer in all three candidate centers. It is the first time that the union is funding supercomputing facilities.

The rest of the funding will go to the building facilities in Italy and Finland, along with smaller supercomputing centers in Bulgaria, Czech Republic, Luxembourg, Portugal, and Slovenia.

MareNostrum was the most powerful supercomputer in Europe when it was built and booted up for the first time in April 2005. More recent versions continue to be used today for scientific research, meteorological simulations and designing medicines.

What is a supercomputer?

A supercomputer is a powerful computer with the architecture, resources, and components enabling it to process large amounts of data at high speeds. The latest generation, made up of tens of thousands of processors, can perform 150 billion computations per second.

The MareNostrum 4 supercomputer, which is currently the BSC's quickest, can perform closer to 10 billion per second, so the European project will represent a significant qualitative leap.

Anthony Elliott

Julie Hare

When you’re lost, Siri can be your best friend. But if she can’t retrieve the right address from your contacts, she can drive you crazy.

And so it is with the legion of virtual personal assistants that are entering our lives. From Amazon’s Alexa to Google’s Home, people are busy talking to intelligent machines as never before.

It’s estimated that more than 60% of internet traffic is now generated by machine-to-machine, and person-to-machine, communication. IT advisory firm Gartner has predicted that by 2020 the average person will be having more conversations with robots than with their partner. (Sometimes we don’t even know we are doing it).

And just as texting changed written communication, talking bots could change the way we communicate with each other.

Talk is social

The late sociologist Diedre Boden wrote that human sociability is created through “talk, talk, talk and more talk”.

Talking person-to-person is not only how we exchange information, but also how we used to carry out many tasks, such as ordering pizzas, booking plane tickets and confirming meetings. And it’s these tasks that we are increasingly subcontracting to robots.

When we communicate face-to-face there is an expectation of mutual attentiveness, but these norms could be wholly deconstructed if we were to have the majority of our conversations with non-humans.

Unlike face-to-face talk, chatbots do not require us put effort into making the conversation polite or interesting. We don’t need to be charming, amusing, or assert our intelligence.

Bots don’t need to like us, even if we have a need to be liked. In fact, this would wildly complicate matters. A machine will simply extract the information it needs to create an appropriate response.