Applied Artificial Intelligence: An Introduction For Business Leaders

by Mariya Yao

Systems That Evolve.

How we build today’s Systems That Learn, Systems That Create, and Systems That Relate will affect how we build tomorrow’s Systems That Master and Systems That Evolve.

The fields of pathology and radiology, both of which rely largely on trained human eyes to spot anomalies, are being revolutionized by advancements in computer vision.

We cannot blindly trust the output of automated systems without vetting the accuracy of both the input data and the decision-making process itself.

While price discrimination based on race, religion, nationality, or gender is illegal in the United States, enforcement of existing law is challenging in e-commerce, where the evidence of differential pricing is obscured by opaque algorithms.

We don’t have to wait for AI to gain sentience and go rogue, because the probability of bad people taking advantage of intelligent automation for evil purposes is 100 percent.

AI can be used for illegal surveillance, propaganda, deception, and social manipulation.

As AI systems become more complex, the likelihood of facing ethical dilemmas also grows.

IEEE, the world’s largest association of technical professionals, published Ethically Aligned Design, a set of standards for the ethical design of artificial intelligence and autonomous systems.(38) The publication lays out the chain of accountability for design and operation.

“Ethics training should be a mandatory part of engineering and computer science education,”

The absence of experiential knowledge means that you cannot solely rely on data and algorithms to tell you which problems need solving.

In most cases, having and using a fantastic machine learning algorithm is less important than deploying a well-designed user experience (UX) for your products.

Thoughtful UX design that delights users will drive up engagement, which in turn increases the interactions you can capture for future data and analysis.

In order to develop “thoughtful UX," you’ll need both strong product development and engineering talent as well as partners who have domain expertise and business acumen.

“Tools are not meant to make our lives easier,” says Patrick Hebron, author of Machine Learning For Designers, “[t]hey are meant to give us leverage so that we can push harder. Tools lift rocks. People build cathedrals.”(40)

Intuition-driven approaches may have worked in a bygone business era when no one had access to data or computing. However, now that software has eaten the world,(41) fortune favors the nerds.

Unless your technology initiatives are driven by clear business goals and viability, you run the risk of using AI aimlessly, like a hammer looking for nails. Deploying AI successfully also requires that your organization be “AI-ready," i.e. have a strong culture of data-driven decision-making and technical experimentation.

promote a healthy work environment in which both humans and machines cooperate and thrive.

AI systems largely handle individual tasks, not whole jobs.

Accenture’s Operations group, which has more than 100,000 employees, initially

calculated that automation would replace 17,000 jobs in their accounts payable and marketing operations. However, headcount actually grew as employees moved to more strategic advisory services that expanded their business lines.(52)

A data science team manager understands how best to deploy the expertise of his team in order to maximize their productivity on a project.

ML engineers build machine learning solutions to solve business and customer problems. These specialized engineers deploy models, manage infrastructure, and run operations related to machine learning projects.

Data scientists collect data, spend most of their time cleaning it, and the rest of their time looking for patterns in the data and building predictive models.

The composition of a machine learning team will change in response to the nature and timing of the project.

While bug fixes and operational maintenance are required after completion, successful software development projects start with relatively clear specifications and product design and are launched when they meet release requirements.

By contrast, machine learning is highly exploratory and experimental,

A background in mathematics and statistics is far more valued in machine learning than in traditional software engineering.

Focus on finding applicants who are particularly excited by the unique problems that you face and the datasets that you own.

Perfect is the enemy of the good.

AI talent tends to evaluate offers on the following areas:

Availability of Data

Quality of Data

Diversity of Problems

Quality of the Team

Impact of Work

Common frameworks include Gap Analysis and SWOT.

Benchmarking helps you to understand where you are and where you want to be.

Don’t limit yourself to data from your own industry. Technology companies have disrupted many traditional business

models, so look beyond the obvious comparisons.

AI Strategy Framework

First, understand the project’s strategic rationale. How does the opportunity fit into your company or department’s overall goals and strategic plan?

Next, consider the opportunity size.

Then, consider the investment level required.

While the next factor, the return on investment (ROI), is never certain, you should estimate an upper and lower bound and a likelihood of success.

The fifth factor to consider is risk.

Also consider the industry risk of your competitors adopting AI for a core

function.

Timeline is the next factor to consider.

Finally, have other business stakeholders bought in?