The conversation is no longer whether artificial intelligence is going to affect our lives or the workplace. Instead, it revolves around a series of questions: When is it going to happen? What will it look like? And will robots replace the entire human workforce…or just most of it?
The answers? AI is happening now. It varies from insights and recommendations to entirely self-driven complex processes. And it will join the workforce, not replace it.
Man + Machine instead of Man Vs. Machine – The Future is Collaborative
In 1997, IBM’s Deep Blue computer defeated reigning world chess champion Garry Kasparov in a high-stakes, man vs. machine chess game. Commentators hailed the victory as proof that AI could match, if not surpass, human intelligence. In 2011, IBM’s Watson competed in the game show Jeopardy! against former champions Ken Jennings and Brad Rutter, and won. The predictions of AI enthusiasts appeared to be coming true: machines were learning to outthink humans.
Yet in between these two landmark cases, another gaming event occurred which may ultimately prove to be the most instructive about the future of AI and its application in our lives. In 2005, the website Playchess.com hosted a “freestyle” chess tournament in which various combinations of humans and computers could compete against each other. Even the most advanced chess computers fell easily to a human player with an average laptop.
In the end, the winner was not the best chess player using the best machine, but two chess amateurs who were particularly skilled at coaching their three computers to find ways to defeat opponents. Success, it turned out, lay not in man or machine alone, but in the hands of individuals who best knew how to optimize the abilities of technology towards a desired end.
In another example, Billionaire Elon Musk’s latest suggestion might just save us from being irrelevant as artificial intelligence (AI) grows more prominent. The Tesla and SpaceX CEO said recently that humans need to merge with machines to become a sort of cyborg. He thinks that over time we will probably see a closer merger of biological intelligence and digital intelligence. That It’s mostly about the bandwidth, the speed of the connection between your brain and the digital version of yourself, particularly output.
Computers can communicate at “a trillion bits per second”, while humans, whose main communication method is typing with their fingers via a mobile device, can do about 10 bits per second. Some high bandwidth interface to the brain will be something that helps achieve a symbiosis between human and machine intelligence and maybe solves the control problem and the usefulness problem. In an age when AI threatens to become widespread, humans would be useless, so there’s a need to merge with machines, according to Musk.
It’s becoming increasingly clear that the most promising applications are not in machines that authentically think like humans, but in AI that complements human endeavors, a field called intelligence augmentation. For context: Scientists distinguish between general AI (machines that possess a wide range of cognitive abilities, as humans do), and narrow AI (machines that can carry out circumscribed, pre-defined tasks). So far, no true application of the former exists, according to Sandy Pentland, one of the founders of the MIT Media Lab. However, there are hundreds of fascinating examples of narrow AI, including those within the field of intelligence augmentation.
Algorithmic Accountability is Mainstream now
We are aware that Algorithm Economy has seen mainstream acceptance, especially in AI. For example, online movie rental service Netflix employs a recommendation engine based on machine learning that predicts which movies users will most enjoy—it accounts for 75% of Netflix usage. Similarly transformative, data-centric technologies are evident in everything from apps like Waze, which helps determine the best driving routes, to programs that help doctors make the best diagnoses in critical cases. And in carrying out such tasks, computers never suffer from mental fatigue or inconsistency due to subjective concerns. By taking care of routinized, time-consuming work, AI frees up people to do the things that machines cannot: plan, improvise, strategize, and decide.
However, Man proposes, but Machine disposes. Society has rules, regulations, and transparency built into its fabric. AI does not. We, our societies, our cultures, need to reinvent the rules of engagement and governance, privacy and human rights, and apply them to our machines.
We all have biases we work around. But when we build them into our machines, biases are amplified with appalling efficiency. For example, when Amazon began launching Prime, offers to join skipped over communities that were predominantly African American. The training data, or the rules, or both, resulted in discriminatory behavior. It wasn’t intentional, but the algorithms did make those decisions.
Who takes responsibility? Most of us have given some thought to the life-and-death decisions the self driving car’s infrastructure may have to make. Someone is coding the decisions into the machine. We need transparency about what the decision will be, what decisions the machines make and why. And we need to decide where responsibility lies.
AI provides Administrative Autonomy while Freeing Humans for Creative Composition
Rather than just providing insights to humans so they can make the most informed decisions and act on them, we’ll see more and more instances of AI autonomously acting on its own insights and carrying out the recommendations that it would normally advise a human to handle. As a result, AI’s future lies in handling a dominant portion — if not all — of the complex processes usually handled by teams of humans and multiple technologies. Depending on the process at hand, this will free humans to focus on higher-level, strategic or creative functions, or those relating to fundamental business decisions rather than execution.
For example of AI working with man comes from the retail industry, where a company called Everseen is integrating its software into security cameras at large retailers’ store locations in order to detect checkout scanning errors that are otherwise undetectable by the human eye. When a product isn’t scanned, Everseen sends an alert with an image of the non-scanned item to stores’ security teams via a smartwatch, tablet or some other mobile device so staff can prevent theft or loss before it happens. The collaboration between AI and human teams could potentially save Everseen’s clients — five of which happen to be among the world’s 10 largest retailers — millions of dollars per year, in an industry where loss of inventory at checkout costs $45.2 billion per year.
Humans Complement AI more than AI Complement Humans at present
However, as AI augment Humans in their various day-to-day and business related tasks, it is also true that Humans augment AI in its learning and evolution. For all the recent advances in AI, human beings remain more adept than machines at distinguishing, say, a tile mosaic from a similar pattern on a blanket. It will be a long way out before machines will be able to do this. Hence, the avenues which call for, human intelligence intensive work, even at micro aspects, are still handled by humans.
For example, engineers at Pinterest constantly create new artificial-intelligence algorithms to help its users find what they’re looking for among billions of pictures of food, products, houses, and other items. Matching search queries with relevant images is crucial to keep users coming back. But until last year, it could take days to test the effectiveness of each new algorithm.
To fine-tune its machine learning and provide better search results faster, Pinterest turned to an unexpected source: human intelligence. It hired crowdsourcing companies such as CrowdFlower to marshal people to quickly do “micro-tasks” such as labeling photos and assessing the quality of search results. In an hour, the workers collectively could test hundreds of search terms to see if results matched well enough.
Pinterest’s experience reveals a sometimes forgotten truth: AI and machine learning depend on people as much as on math. Google’s search engine and ad system use thousands of human “raters” to assess the quality of its AI-driven search results and help identify scam ads. Facebook’s facial recognition software asks people to label their photos to improve accuracy. Deep learning, a branch of AI responsible for recent breakthroughs in speech recognition, language translation, and image analysis, can require extensive human training on hand-picked data sets.
Quantum Thinking Beats Newtonian Thinking – Exploration Instead of Exploitation
It’s better to keep trying fast than to try for the optimum path up front. This is a key strength of continuous learning: machines start out with mostly unsuccessful guessing and eventually get far better at a task than humans. Humans of course use the OODA loop framework to explore new capabilities, but not at the scale or speed of machines.
For this reason, we can now begin to move away from hard wired software, away from Newtonian thinking and toward Quantum thinking. The rules of AI are not rigid, so they support exploration rather than exploitation. Exploitation will continue to degrade as a tool of engagement. Today, if I look at shoes for 30 seconds, I see shoes for days and days, long after I’ve lost interest. That’s exploitation. I think we will increasingly see continuous learning provides the structure for engaging interactions, far exceeding the effectiveness of rules-driven exploitation.
Finally, where is all this leading to ? – Humanism is the new Focus for Civilization
We need a new future of work. If AI obviates human work, despite my hopes for man-machine collaboration, what will people do in order to have not only food and shelter but purpose in life? If not work, what are the avenues to contribute to society, and find meaning? Politicians, lawyers, economists and business executives have been the driving energy in our society, but we need social scientists, psychologists, philosophers, historians, mentors and ethicists to find the path ahead. We need guardrails as well as a roadmap to understand and traverse the ramifications, impacts, and issues. I wish I could envision how this will transpire.
If your company is investing in AI, you need to be aware of these ramifications. You need to promote a path that has the potential to enhance people and culture; that promotes exploration over exploitation. You can strive to be a leader in helping us all find a way through the coming transition.