Tech Emergence – Advocating a More Sustainable Business Culture in an Automated World

Listen to the podcast at Tech Emergence
How does automation influence society today? This is an open-ended question with likely endless answers that can be observed in many different areas of society. As a Writer, Speaker, and Professor in Media Theory and Economics, Douglas Rushkoff has made it his livelihood to examine the impacts of automation in our evolving digital society. In this episode, we speak about his ‘disappointment’ in how automation has been used by many industries without regard for employees’ long-term well being, and how a cultural shift in business’ priorities may be what’s needed to make automation beneficial for the majority.

There are many directions that one could take in answering how automation has impacted society in the last decade, but it’s undeniable that automation in business, especially in a largely capital-driven economy, has far and wide-reaching ramifications. “I have to look at automation within the context of the business landscape itself, and most of the automation you’re going to see funded is by companies and people looking to save money, extract more value out of the world and convert it into their share price.”

When looking at automation in industry, Rushkoff identifies robots and computers that automate human tasks and labor and suggests that the use of these tools, for the most part, is shaped by our market priorities; those companies that have enough money and time to automate processes are usually working for those (owners and shareholders) who are looking to extract more value in less time. “I’ve been a little disappointed…there’s a kind of a reversal  going on where I’m seeing people doing more automated tasks and machines getting to do more interesting ones,” says Douglas.

Opportunities for profit weave into the automated behaviors that spill over into our daily and routinized actions. Especially in more urban and suburban areas, it’s a rare occurrence to not see people constantly swiping through their mobile phones, with instant access to emails, Facebook and other social media platforms, and any number of apps. “Kids went to Stanford and figured out what’s the most Pavlovian effect, what’s the best behavior modification that we can put into an interface to induce people to check it again and again…what can we do to make these devices more and more compelling in order to get more and ore more predictable, repeatable behavior out of people,” says Rushkoff.

The constant search queries and engagement with Facebook and similar interfaces feed back into marketing. Our Facebook news feeds, for example, are filled withalgorithmically-created ads based on a person’s statistical profile, with the goal of getting individuals to behave more consistently with their target demographic group. If a company knows that 80% of a targeted demographic group might go on a diet next week, they might want to leverage this statistic by filling your newsfeed with messages or ads promote products or services for eating the ‘right’ foods.

Rushkoff again makes the argument that in many areas, people are the ones becoming more automated. “Look at,say, Amazon Turks; who do the most repetitive, boring tasks? It’s the people for 2 cents a pop finding the number in the frame; the ones (tasks) too boring for computers to do, they give it to people to do.” In general, he doesn’t see companies trying to make life better for people; rather, they’re goal is to reduce short-term expenditure, often at the expense of long-term innovation in order to convert real-time value into share price.

While there are strong incentives to eliminate wasted time, there does appear to be a potential contention working against well being and creative experiences. Douglas believes it really becomes a question about the reasons behind our programming. Goldman Sachs, Morgan Stanley or any of these stock market investing companies are turning to algorithms that can do high-speed training and predict market trends, with up to 80 percent to 90 percent of transactions now automated.

That’s fine, says Douglas, but what are these bots really programmed to do? In reality, he says, it seems to exploit infinitesimal errors in arbitrage opportunities in the market. “They’re not there to promote the distribution of capital to businesses that need it; the original function of the stock market, which was to allocate capital, is now being changed because the majority of transactions are automated,” says Rushkoff.

He emphasizes that, like any other technology, it’s not automation in and of itself that is the problem. “The priorities of the particular automation that you’re using, the way in which your automation has been optimize, can end up having extreme effects on the underlying system that you’re automating,” says Douglas.

Can Automation Change an Hourly-Wage-Driven Culture?

Can we get businesses thinking more about the common good than maximum profit without legislation? Are there examples of businesses that are doing so today?

“It’s hardest for public companies because of the fiduciary responsibility and shareholders,” admits Douglas. In a digital age, it’s a shift from the industrial age era to think about optimizing not for the growth of capital, but for the velocity of money, getting it circulating in better ways. One way to better control long-term interests is for companies to go private or change their corporate structure, focusing on long-term benefits beyond short-term share price. “It’s also a matter of realizing that a lot of things we’re doing don’t need to be scaled up, a lot of things do tend to work better at a local level,” says Rushkoff.

The local food movement may be one of the best examples of this phenomenon. If you want to eat healthier food and have less of a negative impact on the environment, then joining a local community supported agriculture (CSA) is a better option than eating mass-produced food that is shipped from across the country or the other side of the world.

For other large and emerging business platforms, Douglas articulates an interesting idea that doesn’t seem to be often discussed. He notes that companies like Uber and Airbnb, which are making significant disruptions to their respective marketplaces, could set an example for dynamic change if they were to let workers or participants share in the long-term company profit.

Drivers are (more or less) doing research and development (R&D) for a company that has already made its plans clear to one day use automated cars. “Rather than simply do R&D for free essentially, driving themselves out of business, if Uber had set aside even 10 percent of the company for its drivers, they could get a share based on how many rides they gave or how many miles they drove, it’s really simple,” explains Rushkoff.

The grand idea is that drivers have a lasting stake in the bigger picture, an investment in both their and the company’s future. If drivers are participating in future profits in a company that they helped build, then they’re much less likely to be upset when their job gets automated, says Rushkoff. “If you’re going to use people to help teach the machines that are going to one day replace them, all you have to do is let them participate; otherwise, how will we get to a place where we can just sit and drink ice tea where the robots till the fields? We can’t.”

A large part of this shift is conceptual, and requires evolving an idea that has become a culturally-engrained norm. “We need to stop thinking of them as ‘employees’ earning wages over time,” says Douglas. The hourly wage culture began with the industrial age; before that, the majority of people were farmers or craftspeople, they owned what they did. It’s really been only in the past 200 years or so that we’ve started to pay people for their time, explains Rushkoff.

The concept of paying people for time poses a problem as we continue to create machines and increase computing power, shrinking the amount of time that it takes to complete many tasks in various industries, and as a result inevitably eliminating certain segments of the workforce. If we can change the structure of earning wages, suggests Douglas, and workers are made stakeholders in large companies with guaranteed long-term incomes, then the eventual onset of automation is much more likely to be welcomed instead of feared.

Facebooktwittergoogle_plusredditlinkedinmailFacebooktwittergoogle_plusredditlinkedinmail