The End of Artificial Artificial Intelligence

Prepare for a Future of Machine Learning

Intelligence is the ability to adapt to change. —Stephen Hawkings

In 2005, Amazon released the Mechanical Turk, a tool advertised as “artificial artificial intelligence.” The Mechanical Turk was built as a platform to outsource small pieces of work, called human intelligence tasks (HITs), to people worldwide who would do the typically small pieces of work. Since then, it has been used primarily to scale the work that is difficult for computers to accomplish, such as reviewing the quality of written content or images.1

However, AI and machine learning are progressing at a rapid clip—exponentially, if Elon Musk is to be believed. Regardless of its exact evolutionary rate, AI and machine learning can tackle an increasingly wide array of tasks, including those often performed by Mechanical Turk. The potential power and impact are such that Jeff Bezos included a special warning (or encouragement?) in his 2017 letter to shareholders, in which he advised his audience to “embrace external trends.” “We’re in the middle of an obvious one right now: machine learning and artificial intelligence”2 he warned. When Bezos takes the time to deliver a specific warning, I’d recommend sitting up and paying attention.

World’s Best Marketing & Naming

I admire great marketing and product naming—names enticing the market to eagerly purchase and adopt the product or capability. Great naming almost dares the customer to be stupid for not purchasing. “Agile” is an example. Who would say “no” to being agile?? “Artificial intelligence” is another example. Yes, computers that can “think and reason.” Of course, I want it! Of course, I’m scared by it!

Spoiler Alert - computers can’t think, and there is no “intelligence” in AI. It’s just great marketing!

Artificial intelligence is any system that perceives its environment and takes actions that maximize its chance of achieving its goals - Wikipedia

Artificial intelligence is a model (an algorithm) optimizing to find the “best fit” answer based on the data provided. AI gives the best probabilistic answer to a question or situation. The recent innovations are more to do with the availability of massive compute and storage capabilities (aka “the cloud”), enabling the training of models as well as an endless number of situations and use cases where the models can be deployed — everything from image recognition to natural language processing.

Process Engineering and Definition

I studied industrial engineering in college. Much of the curriculum focused on evaluating and designing the effective flow of work across an organization and applying technology to work design. My early career was spent reengineering and improving business processes — from machine floor control and manufacturing systems to loan approval processes. The golden rule I learned early was to “simplify, integrate, automate” — in that order. This is often overlooked but “simplify” is the critical “work design” phase, largely setting up success for the next two elements. First, define and simplify jobs and tasks. Second, integrate information flow and the interfaces between jobs and tasks. Finally, work to automate the activity.

The vital work organizations can do is to define and simplify how work gets done. Create the Standard Operating Procedure for work in the organization focusing on repetitive, non-creative work. It may seem not very interesting, but this is vital in delivering both benefits today as well as identifying opportunities for integrating and automating.

Defining your processes, functions, and tasks in a deliberate and granular manner, figuring out how to make these processes into services, creating the rules and formulas for the work and decisions. Great building blocks. Understanding your principles, how you make decisions, and the patterns of your logic? Essential. These types of deliberate engineering and introspection are the bedrock of what algorithms need to automate a process.

I’m currently working on a digital agent initiative, and guess what I’m doing? Documenting and scripting the details of how repetitive analyst work happens. Working to understand both the steps, as well as the judgment utilized to make decisions. Our goal is to create a method where white-collar work is modeled and digital agents created to automate and scale these tasks, but simplifying and standardizing the work is critical to improving the digital agents.


“Everything happens over and over again,” explained Ray Dalio, founder of Bridgewater Associates. “Principles are a way of looking at things so that everything is viewed as ‘another one of these,’ and when another one of those comes along, how do I deal with that successfully?”

Dalio built a decision-making system by writing down the criteria of every issue he encountered. This system allowed him to characterize issues, develop criteria, and easily identify the signal from the noise. In addition, he could synchronize with others and convert many of these issues to algorithms.3

In their 2018 Artificial Intelligence Innovation Report, Deloitte characterized the future of AI in executive decision-making as a“partnership,”—one in which humans define the issues and have the final say on the best answer for their business, while AI analyzes terabytes of data to provide a basis for the decision.4

Dalio likens the perfect relationship between human and machine to playing chess side-by-side with a computer. “So, you make the move, it makes a move,” he said. “You compare your moves, and you think about them, and then you refine them.”Needless to say, Dalio continued, it can be not easy to understand cause and effect in a complex, black-box model.

We can leverage his approach to take advantage of machine learning for our core management approaches. Specifically, we can adopt his clarity and meticulous attention to detail regarding thinking through the patterns in our businesses, creating rules to manage them, writing them down so others can use and improve on them, and making computer models from them.


In hindsight, it’s easy to identify the tidal waves of technological progress—the printing press, the electric light, the automobile, the transistor—that drove business and society into entirely new eras. These inventions had mostly positive impacts on society, but their wholesale social applications and adoptions were not without fear and lessons learned.

The next tidal wave of progress will be defined by artificial intelligence applied to both manual blue-collar work and white-collar activities. The new era of work will have people achieving new productivity levels through digital agents working at their request and on their behalf. The productivity of companies needs to be dramatically improved if we are to stay competitive in the future. Be prepared!



About The Digital Leader Newsletter

This is a newsletter for change agents, strategists, and innovators.  The Digital Leader Newsletter is a weekly coaching session focusing on customer-centricity, innovation, and strategy. We deliver practical theory, examples, tools, and techniques to help you build better strategies, better plans, better solutions — but most of all, to think and communicate better

Share The Digital Leader Newsletter -- By John Rossman

Leave a comment



Jeff Bezos, “2016 Letter to Shareholders,” Amazon dayone blog, April 17, 2017,


Ray Dalio, Alex Rampell, and Sonal Chokshi, “Principles and Algorithms for Work and Life,” a16z Podcast, April 21, 2018,


Deloitte, Artificial Intelligence Innovation Report 2018, Deloitte.pdf