When you buy through affiliate links in our content, we may earn a commission at no extra cost to you. Learn 
To say that artificial intelligence (AI) is the next step in enterprise would be an understatement.
AI has already become a reality for many industries, including, but not limited to:
- Health care
- Insurance
- Oil and Gas
- Agriculture
- Publishing and Media
- Architecture
- Hospitality
- Finance
- Customer Service
In other words, the so-called “AI revolution” is already here. Moreover, it’s growing in strength and popularity.
Want Access to the Free PDF Version of this Guide? Download the Techopedia White Paper Here.
According to McKinsey and Company, by 2030, 70% of companies will have adopted at least one kind of AI technology. The expansion of AI also stands to have a significant impact on the world’s economy and job-force.
The hype around AI has also led to a rapid increase in companies investing in AI and big data out of fear of being left behind. According to New Vantage Partners’ 2019 Big Data and AI Survey, 88% of Fortune 1000 companies feel an urgency to invest in big data and AI, with 55% of these companies spending more than $50 million dollars on these investments.
McKinsey and Company also predict that AI technologies could lead to a performance gap between companies that fully absorb AI tools across their enterprises over the next five years compared to those that do not by 2030.
But it’s not just about a company’s fear of being left in the dust; 84% of global business organizations believe that AI will give them a competitive advantage, according to MITSloan Management Review.
How? Let’s start with Chatbots. Innovation Enterprise believes chatbots will power 85% of customer service by 2020…yeah, that’s soon.
And of course, revenue and economic gains are major factors in why companies are realizing AI’s worth. PwC believes AI could contribute up to $15.7 trillion to the global economy in 2030, while Tractica predicts the AI software market to reach $118.6 billion in annual worldwide revenue by 2025.
And when Element AI revealed in its 2019 Global Talent Report that the number of people claiming to have the educational and skills profiles to qualify as an AI expert rose by 66% between 2017 and 2018, it’s easy to understand where the job landscape in tech is heading.
While it is well known that AI is the next step forward, myths and misconceptions about AI and its processes still run rampant. It lead us to an important question:
How well do those in charge of making decisions about AI projects actually understand AI?
So, we created a two-part survey and quiz to help us examine how well industry executives comprehend AI and machine learning (ML)
Before we begin, take a look at how this infographic celebrates over six decades of historic migration of software development practices from the waterfall to new DevOps and agile methodologies.
Share this Image On Your Site
Where Did the Term “Software” Come From?
The term “software” was coined in 1953 by 19-year-old Paul Niquette who programmed the Standards Western Automatic Computer (SWAC) at UCLA, one of only 16 digital computers in the whole United States.
When I first said “software” out loud, people around me said, “Huh?” From the very beginning I found the word too informal to write and often embarrassing to say. Nevertheless, with smirking trepidation I did occasionally feature the word “software” in speeches and lectures and media interviews
The Waterfall Model
As computers spread, Dr. Winston Royce’s Waterfall Model (1956) instructed companies on how to produce software in the shortest, most effective manner.
This logical, hierarchical system was introduced in a journal article by Bell and Thayer in 1976 and later standardized by the US Department of Defense in 1985 for its DoD software development providers.
The Six Stages of the Waterfall Model
- Preliminary Design
- Detailed Design
- Development
- Unit Testing
- Integration
- Testing
Still used today, the model is best suited for larger projects and organizations that could benefit from its stringent stages and deadlines.
The Spiral Model
In 1986, Barry Boehm combined the Waterfall model with iterative procedures in a system he called the Spiral Model.
Each of the four phases of the Model begins with a design goal and ends with the client reviewing the results. Each phase, too, has its own particular task.
The Spiral Model best suits large and complex unpredictable projects.
Approaching software development from another tangent, Frederick P. Brooks noted in his 1975 book: The Mythical Man-Month, that “adding more men lengthens, not shortens, the schedule.”
Brooks followed that up with an article called No Silver Bullet—Essence and Accident in Software Engineering that argues that since no single software has ever been completely error-free, we need software development methods for simple and reliable software.
The Scrum Framework
In that same year, 1986, Takeuchi and Nonaka introduced the term “Scrum” that, they said, was an approach to organizational knowledge creation:
“In today’s fast-paced, fiercely competitive world of commercial new product development, speed and flexibility are essential.”
The Scrum framework recommended that software development move from a “rugby” to a “relay” approach. The relay software development approach uses the traditional/waterfall method where one group passes on the process to the next.
The rugby model, in contrast, “passes the ball” within the team as it moves as a unit up the field. The last is faster and makes for more autonomy and stability.
The SCRUM Development Process
Three years later, Schwaber and Sutherland incorporated “Scrum” in their respective software development companies. And in 1995, Schwaber and Shuterland published The SCRUM Development Process that outlined the Scrum approach.
Although Scrum is relatively simple to understand, new users may find Scrum — with its values, roles, events and artifacts — complex
Rational Unified (Software) Process
One year later, IBM’s Rational Software Company created a Rational Unified (Software) Process(RUP) that splits Scrum’s software project life cycle into four phases.
Each phase contains all of the six core development disciplines, while certain processes are given more attention than others.
So far, innovators largely had midsize to large companies in mind. In 1999, Kent Beck authored his book Extreme Programming Explained for small businesses that struggled with vague and changing software development requirements.
The book introduced the Agile revolution for software processing, where the collaborative agility of Scrum was said to be better than the rigidity of the Waterfall framework.
The Agile Manifesto
In 2001, 17 software development practitioners gathered at a ski resort in Utah to ski and relax. In their spare time, they compiled their Agile Manifesto with its four values and 12 principles for more agile software development processes
“Through this work we have come to value: (1) Individuals and interactions over processes and tools; (2) Working software over comprehensive documentation; (3) Customer collaboration over contract negotiation, and (4) Responding to change over following a plan”
Later, Robert “Uncle Bob” Martin added a fifth value called: “Craftsmanship over Crap”, at his Agile 2008 keynote in Toronto.
Lean Software Development
Do you want better, faster, cheaper software development? Mary and Tom Poppendieck identified seven fundamental “lean” principles in their 2003 book Lean Software Development that could be adapted to the world of software development.
Each of these principles already revolutionized manufacturing, logistics and product development. They could be applied to software development value, flow, and program developers.
DevOps
Meanwhile, a frustrated Beligan consultant, called Patrick Debois, developed and promoted the term DevOps through his presentations.
DevOps simply means the cross department integration between Development, the department creating the code, and Operations, the department using that code.
Of course, it’s far more complex than that, and its ramifications on software development and deployment were immense. By unifying development and operations, DevOps has created the opportunity for everyone to participate.
Kanban
One last influence was David Anderson’s 2010 book Kanban. Less a software methodology like Scrum, Kanban, sourced from Japan, tells you how to continually improve your software features, products or services.
In that way, its methods can be applied as much to Agile models like Scrum as to traditional models like the iterative process.
Where Does that Leave Us Now?
We’ve come a long way from when Paul Niquette coined “software”. First off, the software development ecosystem — now integrated into DevOps — has acquired an array of tools from Dockers to Kubernetes, Puppets, Chef and Ansible.
As to the best method for software production, Agile frameworks have largely replaced traditional models, with these structures helping teams plan, manage and deliver software on time.
The most popular of these are the Scaled Agile Framework (SAFe), Large Scale Scrum (LeSS) that’s a scaled version of a one-team Scrum, and Disciplined Agile Delivery (DaD).
Expect more developments with a transformative future.
The results of our survey supported one clear answer: Business and industry executives do not understand the majority of AI and ML.
These are the leaders in charge of implementing AI projects, the ones responsible for determining how AI and ML will be used and what it will be used for. And that leads to another key question:
How can AI be used effectively if those seeking to use it don’t understand it?
In order for AI and ML to be used to their maximum potential to help streamline enterprise, reduce costs, reduce risk and increase profits, it needs to be implemented with precision by those with realistic expectations.
The drive to incorporate AI and ML into enterprise is there. Of the respondents, 44% had ongoing AI or ML projects, 14% were working towards implementing it, and 17% were not yet working on it, but want to.
Students Scored an Average of 56% on the Quiz, Whereas C-Suites scored 51%
Our survey and quiz showed that myths about AI and ML are prevalent among industry executives. It’s also interesting to note that students scored higher on the AI and ML quiz than C-Suite level executives did, even those engaged in ongoing AI and ML projects at their company.
On the other hand, it’s not surprising to see that students outperformed C-Suites. While classes in AI and ML are now common at all technology-centric universities, that is a recent development. Chances are, when most of these C-Suites were in school, taking courses on AI wasn’t an option.
31% of Respondents Answered That AI and ML Are Two Completely Separate Things
For those familiar with AI and ML, it is well understood that ML is a subset of AI. However, 31% of our respondents answered that AI and ML were two completely separate things. Furthermore, 23% of respondents believed that AI programmed computers have the ability to exercise free will, something that any AI or ML engineer would scoff at. In addition, 47% of our respondents believe that one of the main challenges of AI is that the further AI progresses, the larger the danger of AI programs creating sentient machines that pose a threat to humans.
Read: Data Science or Machine Learning? Here’s How to Spot the Difference
55% of Respondents Believe AI and ML Will Increase Unemployment
While studies have supported that AI is more likely to result in job shifts and redesign, 55% of our respondents answered that they believe AI and ML will increase unemployment rates in the long run.
The results of our survey lead us to believe that in order for AI and ML to succeed, enterprise needs to truly understand what these tools are and how they can help bring about innovation.
What Executives Need to Understand About AI
First thing’s first, C-Suites and executives need to make sure they are investing in AI for the right reason: because they have a specific problem they want to solve.
Investing in AI out of fear of being left behind is bound to lead to disappointment and further misunderstanding of what AI is truly capable of.
Of course, we can probably trace the confusion around AI back to science fiction. While we’ve all probably seen at least one movie with a super-intelligent AI system or robot threatening humans, none of us has seen a movie following the implementation of a successful AI project that helps to simplify and increase the efficacy of a workflow. At least … not yet.
In other words, AI in enterprise does not lead to super-intelligent robots that are out to kill the entire human race.
Now that we’ve gotten that myth cleared up, let’s examine what an AI in enterprise project really looks like. It starts with having a clear vision of what part of your workflow can be solved with AI.
We wanted to help bridge the gap between the AI and ML experts, and those who want to introduce AI into their business. In doing so, we reached out to various experts in AI and asked them what it is they wished C-Suites understood about AI.
As we talked with AI experts, the same phrase came to light over and over again:
AI Starts With Data
It is a common misconception that you can start (and finish) an AI project simply by throwing a mass amount of data at it. This is far from the case.
Only 48% of respondents of our AI/ML quiz knew that a smaller, more specific amount of data that is highly relevant to the question being asked is better than a large amount of data, only some of which is relevant to the question being asked.
This lack of understanding goes hand and hand with the myth that AI and ML can solve any problem. As we previously explored in our infographic “Choose Your AI in Business Adventure,” many industry representatives misjudge what it means to implement an AI project. AI projects don’t need to be giant moonshots. Instigating an AI project doesn’t mean replacing human workers with robots or building drones to do extra work.
The best use-cases for AI and ML projects will reduce costs, reduce risk and/or improve profits. More often than not, the best results are seen from implementing AI to handle the small, repetitive tasks that businesses do on a daily basis.
Of course, finding these projects is not always the easiest task for C-suite and upper-level executives. Narrowing down an AI or ML project requires having an in-depth knowledge of a company’s workflow.
It is also important to understand that implementing an AI project is not simply flipping a switch to create results. AI projects need to be carefully planned and will take time to show results.
However, if the correct plans are made and the cooperation between the AI experts and the executives is there, AI is still an incredible step forward for enterprise.
While it would be great for executives to have a thorough understanding of all aspects of AI and ML, it’s a dream that likely will not happen, and fortunately, it’s not necessary for successful implementation of AI into business.
All we need is a more general understanding, and one which we believe the following guide will help with:
The Ultimate Guide to Applying AI in Business
1. Ask yourself and your company “What’s our problem?”
What do you want to predict? This needs to be specific. Simply wanting to grow your business is not the right strategy to implementing AI.
2. Examine your business’s workflow and START SMALL.
Where in your business would AI be the most effective? Don’t go for the moonshot, look for the simple, repetitive tasks that AI and ML can help accomplish.
3. Examine your business’s data flow.
Now that you have a problem at hand, look at your data. Do you have the data needed to answer this problem? Chances are, examining your data will result in you performing a deep clean of your data and narrowing it down to the data that is truly relevant to the task at hand. This may mean bringing in an educated data science team that can help you analyze your data and determine how useful it is for the project you want to start. Remember that a smaller, more specific data set is better than a large, less relevant data set!
4. Determine the return on investment (ROI) for your project.
What measurables will be most affected by your AI/ML project? How will you track these? Have a clear expectation as to when you will start to see results.
5. Determine whether you have the in-house talent to start your own AI project.
While AI skills and capabilities are becoming a more common, sought-after skill, it is important to not underestimate the work and time that goes into an AI or ML project.
6. Find the right external AI/ML partner.
With the boom in AI and ML, there are many companies out there that specialize in helping your company find its AI potential. But not all AI companies are created equal. Find the right partner with the experience, capabilities and resources to help bring your project to light.
7. Be patient, yet agile.
This is a big one to remember. AI projects are not sprints, and you have to crawl before you can run. AI will impact many aspects of your business and the models will require learning time and practice in order to produce results. Sometimes moving forward will require rethinking and re-examining of your initial problem.
