Five Reasons JARVIS Is Not a Job Killer

Manoj Saxena, Executive Chairman, @manojsaxena

One of the main themes at last month’s World Economic Forum in Davos was the “Fourth Industrial Revolution,” or the ongoing impact of smart machines on the global economy. Advances in artificial intelligence and robotics will help us address some of the world’s most difficult problems, like cancer and climate change. At the same time smart machines will be globally disruptive, forcing millions of people out of their jobs and requiring them to retrain.

The debate at WEF reflects the growing understanding that smart machines can outperform and potentially replace humans in a huge range of tasks, from reading a brain scan to driving a taxi. A WEF report released in January 2016 finds that increased AI and automation will cost the world’s top 15 economies 7.1 million jobs over the next five years, while creating only 2 million new ones. Several leaders including Satya Nadella, chief executive officer of Microsoft, stressed the importance of education and training to help humans pick up new skills.

We can all agree on the importance of education and training to prepare us for the jobs of the future. But what’s less understood is exactly how smart machines will enhance our work, rather than take it away. Perhaps one reason is the “Uncanny Valley,” the computer science term for our instinctive discomfort with artificial humans. The Uncanny Valley is often used as a plot device in Hollywood films, where the machines either betray us (Ex Machina and Transcendence), or their not-quite-human nature is used for comic relief (Star Wars). We just have a hard time imagining a future in which humans and machines work seamlessly and effectively together.

Here at CognitiveScale, a provider of machine-intelligence powered cognitive computers for major industries like retail, healthcare, and financial services, we firmly believe that future is real, and it’s coming sooner than expected. We call that future “augmented intelligence.”

Augmented intelligence encompasses machine learning and cognitive computing, where technology serves to augment and amplify human intelligence, not steal our jobs. In the new AI, it’s not man versus machine. Rather, it’s man and machine working alongside each other to improve the way we work, live, and play.

Here are five reasons robots won’t take over our jobs and why we’re going to need augmented intelligence more than ever.

Putting AI to Work For Us, Not Against Us

1. Genuine machine intelligence requires a lot of training and real-life experience.

Most of the information needed to replace a human’s ability to solve hard problems result from years of training, applying, collaborating, and learning — by humans. Educating smart machines requires running experiments and training machine learning models on massive amounts of data, much of which isn’t documented or clearly and easily identified.

This, in turn, this will create an entirely new family of jobs to train computers to learn new fields and provide sophisticated services to help drive better decision making, improve customer and user engagement, and lead to more productive work environments.

2. You can’t model and emulate what you don’t understand.

The human brain is the most complex and least understood part of the body. While digital computers are capable of emulating the behavior of other digital computers because they function in precisely defined ways, this is not the case with human intelligence. Neurons in the human brain are complex analog systems whose behavior can’t be modeled on a computer the way that digital circuits can. The slightest imprecision can lead to a wildly inaccurate model of the brain.

As advances in AI continue, the development of intelligent machines will remain a slow and gradual process, especially since we are light years away from fully understanding how brains work. Even when we do possess a greater understanding of the brain, computers with superhuman intelligence will continue to be extremely dependent on humans. This is because human consciousness, and the brain’s unpredictable, nonlinear interactions among billions of cells, can’t be completely replicated in silicon.

3. The unabated data explosion will create new and different jobs.

Two years ago, 90 percent of today’s data didn’t even exist and the data deluge shows no signs of slowing. Since the arrival of “big data,” we’ve already seen the creation of new roles, including “data scientist.” The new role was necessary since so much of today’s data is unstructured (text, social, images, etc.), which makes it difficult to put into context.

While AI can tame big data by giving businesses insight, it’s the employees’ knowledge about non-data factors, such as the top three business goals of the company, and how they apply to the data, that brings value in a way that a machine cannot. Make no mistake — we’ll continue to need analytical expertise to make sense of the mounds of data. In fact, McKinsey & Company estimates that the US is currently facing a shortage of approximately 165,000 people with analytical expertise and 1.5 million managers and analysts with skills to make decisions based on their data.

4. AI may take away old tasks, but it also creates new ones.

AI systems usually replace individual tasks, not entire jobs. Some aspects of legal work, for example, may be automated, but selling work, communicating with clients, and interpreting the results of AI analyses will still require human lawyers. And almost as rapidly as it automates old tasks, AI leads to new ones. In law, there are now “predictive coding” specialists who work with machines on that task. In other industries, machine augmented processes and applications make it possible for companies to produce higher value sophisticated goods and services — smart thermostats, smart watches, smart jet engines. These products open up all kinds of new employment opportunities such as customer support and user experience design in companies and countries in which these goods and services are produced.

5. We’ll still want humans to make most life-and-death decisions.

Health care is one of the most important application domains for AI. More than one in every two adults will have a chronic condition by 2020, costing $2.4 trillion annually. Yet the most common chronic conditions — heart disease, stroke, cancer, diabetes, obesity, and arthritis — are the most preventable. Through advances in AI, as much as 90 percent of healthcare can be done outside the clinical environment.

But our experience thus far is that patients, providers, and payers in health care are all more comfortable with doctors and nurses making the final decision. In radiology, for example, computers can detect breast lesions and colon polyps, but these automated scans are only a “second set of eyes” for human radiologists. When diagnoses and treatment options are automated, patients and caregivers prefer a system that provides recommendations with probabilities, but leaves room for insight and interpretation by experienced humans.

There is huge potential for increasing the productivity and lowering the cost of health care through AI. But anyone who thinks that doctors and nurses will disappear should have his head examined.

Leave the Robots to the Movies

Robots are everywhere. There’s a glut of them. We all enjoy digging into a bucket of popcorn as we escape to the imagination of Hollywood and its depiction of man versus machine. Humans becoming slaves to robots makes for fun journalism as well, as word from the labs keep reminding us of the latest advances in robot technology. But if we step back from the hype and depictions of a doomsday future, the paramount issue today is all about the way we can use data and our own intelligence to determine the role that smart machines will play in our lives. The future isn’t inevitable, it’s what we make of it.


Manoj Saxena is the founding managing director of The Entrepreneur’s Fund IV, a seed fund focused exclusively on the Cognitive Computing space. He is the Executive Chairman of CognitiveScale, Chairman of the Board at SparkCognition, and also serves as a special advisor to IBM senior leadership. Read more about Manoj here.

#CognitiveComputing #AI #MachineIntelligence #MachineLearning #CognitiveEra #ArtificialIntelligence