Skip to main content

The One Thing Software Can't Eat

The future of software is driven by human brilliance, with AI in the back seat.

Signature of Gordon Ritter

The future of software is driven by human brilliance, with AI in the back seat.

At the beginning of the last decade, Marc Andreessen declared that software is eating the world. He predicted that “software companies are poised to take over large swathes of the economy.” And by and large, he was correct. Software took the world by storm. It enabled a more efficient workplace and allowed enterprises to scale at such a rapid pace that the new generation cloud-based companies that spun out of this era are now some of the largest and most influential in the world. 

Andreessen’s famous statement, now nearly a decade ago, accurately captured the digital transformation almost every industry has seen. However, in the last few years, we’re starting to see some course corrections to the eye-watering valuations and the growth-at-all-costs mentality. 

One thing has become increasingly clear: software can’t eat people.

As we kick off a new year and new decade, it is time to evaluate the long-term impact of what we’ve been so rapidly building. Companies have been in fierce competition to leverage machine learning and artificial intelligence technology to scale bigger, better, and faster. Software can be reprogrammed faster than humans. Businesses only focused on growth-at-all-costs, if we are not careful, could degrade human talents and remove them from the equation entirely.

This is not to blame any one company or industry; It has been a collective effort. As companies grow, far too often, a Borg groupthink takes over with an insatiable appetite to keep growing, sometimes at the expense of doing the right thing. We’ve seen Uber and Amazon take Andreessen’s message too far, too fast. These companies were thought to have strong business models, but at the core, their vision was to automate the human role across their respective industries. Now that the doors have been blown open, it’s clear they’ve relied on treating their workers poorly — denying them fair wages, hours, and dignity — and that their skills have been only a temporary solution in order to harvest the data needed to fully automate their roles. 

People can only be pushed so far before they start fighting back. We’re just beginning to see the blowback. Not everything can, nor should, be software-ified.

Paying the price

Investors are attracted to software businesses because of their ability to scale, as opposed to labor-intensive businesses that require more workers to effectively grow. Yet many of the recent so-called software-based businesses have models that have human labor at their cores. Where they went wrong was, as they grew, they didn’t take care of their people. For a while, that worked just fine because the world was excited about the new opportunities they brought. Private funders pumped money in to support all aspects of the new disrupters. But it has become more apparent that these companies never intended for their employees to be anything more than a cog or a harvester of data. With this shakeout we are seeing more people shed light on labor issues, poor culture and mistreatment of employees.

That’s what Uber has come to realize. Driving is a complex activity that won’t be taken over by software anytime soon; in fact auto and tech industry experts predict it will take at least 12+ years. But the company has acted as if people driving on its behalf were just a temporary solution until it figured out self-driving technology. It is widely known now that Uber “treats their drivers like crap,” a message that has caused tangible consequences. When it hit the public market, the $120 billion valuation target has continued to dip to half that, and the situation will likely get worse before their plan for autonomous vehicles can relieve it, if at all.

Another hard lesson, though more about “bricks” than “biology,” was the implosion of WeWork. The self-proclaimed technology company operated under the pretense it removed the capital intensive process of buying and managing buildings via software. Perhaps it was visionary, but the reality is that physical spaces don’t change their cost structures because of software. And under the hood, WeWork’s model relied on employees to be the cogs creating this “disruptive” new experience.

Humanity strikes back

We don’t get to defy gravity with software. The most amazing software in the world may be able to disrupt certain industries, but human beings and the structures we work in cannot be rewritten in code. Software’s appetite does have a limit. AI and software can accelerate at a rapid pace, and human minds are not tuned to keep up with these algorithms. We’re simply on a different curve. Companies should avoid software to harvest our human capabilities (such as using people to help autonomous vehicles identify stop signs and pedestrians carrying bikes) without also making us better. Forcing humans to serve software, as opposed to the other way around, will cause many people to be thrown off the boat, or, as we’re seeing now, revolt.

California has attempted to fix part of this issue with AB-5. The bill would reclassify the on-demand workforce, such as Uber and Lyft drivers who have been leveraged as independent contractors (much to the companies’ benefit), to full-time employees with benefits. While not without controversy, the bill shows what happens when businesses ignore the needs of people as they serve their software.

These issues are also one reason the presidential campaign of Andrew Yang has exceeded all expectations. Research, reports, and surveys paint a scary picture of automation displacing workers and many people potentially in the line of fire see Yang’s proposal of universal basic income as the solution. 

To be sure, not every company that allows software to feast on human capabilities has faced repercussions. Amazon pushes for efficiency to a fault. There is no shortage of whistleblowers on the unacceptable working conditions in their warehouses or shipping departments, and even reports of deaths by third parties attempting to meet its unreasonable demands. Several years ago the New York Times reported that the company “rejects many of the popular management bromides that other corporations at least pay lip service to” and “is conducting a little-known experiment in how far it can push white-collar workers, redrawing the boundaries of what is acceptable.” 

Unfortunately, Amazon has so far escaped culpability. While Uber has targeted a single industry with a unified set of workers (drivers), Amazon has built its business on the backs of hundreds of thousands of diverse smaller players—it has divided and conquered. Now, because of its immense size, Amazon has mostly stayed protected from outside intervention.

How people and software can work together

No one should accept the fate of becoming a human cog to technology. There are better ways for humans and software to coexist, and businesses should use software to help humans evolve faster. 

At Emergence, we believe in the power of Coaching Networks, a new breed of companies that builds software to help unlock our full potential, and can scale without consuming people in the process. A company that exemplifies this human/machine collaboration is Textio, an AI-augmented writing platform. Textio is not a crutch, it does not do the writing for writers. It calls upon the multitude of shared writers on its network, discovers patterns and suggestions, and lets the user refine their skills the more they use it. is another example of software that acts as an on-the-job coach, in this case  for verbal communications. It integrates with communications infrastructure like  Zoom, adding a layer of “conversational intelligence.” Chorus isn’t trying to harvest and replace human communications—in fact, it learns unique habits of voice-based human communications and shares those habits with others to help them learn and grow. Retail Zipline bridges the gap of physical labor and software, helping retailers maximize the context-aware brilliance of workers to improve the in-store experience.

In all of these cases, employees are at the center of the value of this new type of software. It is not harvesting users' behaviors to find ways to replace them, but instead, Coaching Networks are designed to gather the constantly-evolving brilliance of workers in all industries and share it widely to create a continuous human learning loop. 

What many leave out of the conversation is how bleak, even dangerous, the world would be without the fundamental elements that make us humans: judgments based on context, empathy and values. We thrive on being able to tap into our creativity and solve complex challenges with our critical thinking. And as we’ve seen depicted in many fictional futuristic plotlines, when these elements are threatened and humans are reduced to automatons, we will fight back.

It’s clear that if software tries to eat people, it’ll find that it has bitten off more than it can chew.

Enjoying this article?

Sign up to gain access to our thought leadership and have future articles delivered directly to your email.