Monday, May 22, 2023

Artificial Intelligence is Hungry and Agile is Ready for It


In my blog last week, I argued that the agile community needs to shift its focus from transforming organizations to delivering agility to customers. I promised to provide additional details and explain why this shift is necessary. Today, I want to concentrate on the world-consuming qualities of software development and the changes its bratty cousin artificial intelligence has in store for us. 

Twelve years ago, March Anderson, one of the founders of Netscape, which helped create the internet economy, authored a prophetic article in the Wall Street Journal called "Software is Eating the World." Anderson argued that software was becoming more dominant in managing the global economy. Corporations must behave like software companies instead of traditional businesses to remain profitable. It sounded crazy and like much of the hyperbole surrounding the technology business in the giddy and stupid days surrounding the first dot.com bubble.

It's been years since Anderson's prediction, and it continues to prove itself true. Just take a look at Southwest Airlines, Volkswagen, and Boeing. Inclement weather forced Southwest Airlines to ground flights due to faulty software. Volkswagen found itself in a scandal where special software deliberately altered diesel emissions to falsely show the vehicles were better for the environment than they were. And the software in Boeing 737 jets was directly related to fatal crashes. It is only the tip of the iceberg, however. In our daily lives, algorithms, and software play a significant role, from streaming movie services to grocery shopping. Clearly, Anderson's prediction is more relevant today than ever - software is playing an increasingly important role in the world.

Unfortunately, my experience has shown me that many business organizations do not understand the importance of software until it is too late. Often it is when a competitor can do things at a lower cost or when software breaks when business leaders understand the crucial nature of software. It also became evident that how they were doing business was not compatible with a global economy. It is a harsh lesson, and I continue to see it taught daily in the board rooms and cubicles of modern business. 

The solution to this challenge is to change how we do business from a command and control emphasis to a more collaborative one where ideas blend to help deliver value to customers. It is very alien to people accustomed to giving orders and expecting obedience. Creating software is collaborative, data-driven, empirical, and exhaustive. If it were easy, everyone would do it, but it takes special skills and a particular temperament to be good at it. Furthermore, less than a fraction of one percent of the world population can build software even though we depend on it for survival. So software is eating the world, growing more ravenous each day. 

Where does artificial Intelligence fit into this picture? It will change office work, and we are already seeing companies like Microsoft and Google attempt to add the benefits of AI into their products. Some of it is mighty cool, and other parts of it are forgettable. I see some of the most powerful use of AI in writing. I leverage tools like Grammarly and Wordtune to ensure good grammar and spelling. Lately, I have been an early adopter of the Bearly AI tool, and it has many possibilities but still needs significant improvements to attract the attention of a more mature audience. 

Despite the attention, like the fawning 60 Minutes segment this year, artificial Intelligence has a long way to go. Image recognition systems with AI are terrible.  These systems can not distinguish between a dog and a piece of fried chicken. Artificial Intelligence can write basic software code, but the programs are simple and rudimentary, so it still requires people to take the complex needs of business people and interpret them into code that will live in production. Finally, we have the problem of bias and hallucinations in Artificial Intelligence, which will be an ongoing problem.

Soon, people who can use AI effectively will be like software developers because they will train the systems which make the world operate. So I see Artificial Intelligence not as a threat but as a tool for business. It can help improve productivity and reduce mistakes, but human values and judgment are critical to success in business and life. 

Agile fits into this worldview because it helps you leverage the technology of software and Artificial Intelligence by measuring results, making corrections, and testing them again in real-world situations. If we are to deliver agility to customers, we should be able to inspect and adapt to change as it happens. Otherwise, we are going to crash and burn horribly.

Until next time. 


No comments:

Post a Comment