But there are a number of issues. The first is that the workers on these platforms earn very low wages. We did a study in which we followed hundreds of Amazon Mechanical Turk workers for several years and found that they were making around $ 2 an hour. It is much less than the American minimum wage. There are people who dedicate their lives to these platforms; it is their main source of income.
And that leads to other problems. These platforms also reduce future employment opportunities, as full-time crowd workers do not have the opportunity to develop their skills – at least not those that are recognized. We have found that many people don’t put their work on these platforms on their CVs. If they say they’ve worked on Amazon Mechanical Turk, most employers won’t even know what it is. Most employers don’t know these are the workers behind our AI.
It is clear that you have a real passion for what you do. How did you end up working on it?
I worked on a research project at Stanford, where I was basically a crowdworker and it exposed me to problems. I helped design a new platform, which looked like Amazon Mechanical Turk but controlled by workers. But I was also a technician at Microsoft. And it opened my eyes to what it’s like to work for a big tech company, too. You become faceless, which is very similar to what mob workers experience. And that really made me want to change my workplace.
You mentioned having done a study. How do we know what these workers are doing and the conditions they face?
I do three things. I interview workers, conduct surveys, and build tools that give me a more quantitative perspective on what’s going on on these platforms. I was able to measure the time that workers invest in performing tasks. I also measure the amount of unpaid work that workers do, like finding jobs or communicating with an employer – things you would get paid for if you had a salary.
You have been invited to give a talk at NeurIPS this week. Why is this something the AI community needs to hear?
Well, they feed their research with the work of these workers. I think it’s very important to understand that a self-driving car or whatever exists because of people not being paid minimum wage. As we think about the future of AI, we need to think about the future of work. It helps to remember that these workers are humans.
Are you saying that companies or researchers are deliberately underpaying?
No that’s not it. I think they might underestimate what they are asking workers to do and how long it will take. But most of the time, they just haven’t thought about the other side of the deal.
Because they only see a platform on the Internet. And it’s cheap.
What do we do about it?
Many things. I help workers get an idea of how long a task can take them. This way they can assess whether a task will be worth it. So I developed an AI plugin for these platforms that helps workers share information and coach each other on tasks that are worth their time and that allow you to develop certain skills. AI learns which type of advice is most effective. It takes in-text comments workers write to themselves and learns which tips lead to better results and promotes it on the platform.
Let’s say workers want to increase their wages. AI identifies what kind of advice or strategy is best suited to help workers do it. For example, it might suggest that you do these types of tasks from these employers, but not those other types of tasks there. Or he’ll tell you not to spend more than five minutes looking for work. The machine learning model is based on the subjective opinion of Amazon Mechanical Turk workers, but I have found that it can still increase workers’ wages and develop their skills.
So it’s about helping workers get the most out of these platforms?
It’s a beginning. But it would be interesting to think about the career ladders. For example, we could guide workers to a number of different tasks that allow them to develop their skills. We can also think of offering other opportunities. Companies that create jobs on these platforms could offer online micro-internships to workers.
And we need to support entrepreneurs. I have developed tools that help people create their own concert markets. Think about these workers: they are very familiar with working on stage, and they might have new ideas on how to handle a rig. The problem is they don’t have the technical skills to create one, so I’m creating a tool that makes setting up a platform a lot like setting up a website template. .
A big part of this is using technology to shift the balance of power.
It’s also about changing the narrative. I recently met a couple of crowd workers that I spoke to and they call themselves tech workers, which – I mean, they’re tech workers in a way because that they fuel our technology. When we talk about crowdworkers, they are usually presented as having these horrible jobs. But it can be helpful to change the way we think about who these people are. It’s just another technological job.