The hype over AI is steady and growing with yearly investments surpassing $10 BN USD. Even as this enables less-than-reputable behavior by some—for just $200 USD, for example, hackers can make 3D masks to confuse Apple’s latest face recognition system—trust in AI remains.
But organizations and leaders at all levels should still take a deeper look at how to integrate AI responsibly.
Of course, there are all sorts of reasons for adopting AI, from sales boosting to processes optimization and cost reduction. We interviewed dozens of experts for AI insight on what every organization should consider. Despite the wide range of applications, these experts indicated a consistent logic to implementation.
Namely, that business-related AI implementation should focus on people, individually and as a team.
In line with that, there are three things that make AI successful in a company: inclusive solutions, relevancy, and empathy. Here’s how leaders can put all three to use.
Next Article
How Machines Learn: Supervised, Unsupervised, and Beyond
Successful AI Startups: A Checklist
How To Leverage Innovation Tech and Transform Your Business
AI and Inclusive Solutions
One of AI’s biggest strengths is its ability to spot abnormalities, but experts agree on the continued importance of humans. This is the key to achieving a market use for AI tools.
Leaders hoping to successfully implement AI need to know their teams’ strengths, weaknesses, and needs—as a whole and individually. Any team adopting AI-powered tools has to be able to face the trials of iteration. After all, AI is unique to past disruptive technologies due to its evolutionary nature—AI tools get better at each iteration.
While this has an obvious benefit, iteration comes with growing pains.
A typical mistake when training AI is feeding it biased data, such as observations of customers that only include white working men with a degree. Algorithms need to be audited to consider diversity—often, that highlights a need for diverse coders.
AI iteration is neither instantaneous nor machine led, meaning humans retain a key role. Our research indicates that AI iteration takes at least two years and lots of hands. Data scientists train each AI generation; salespeople produce insights for practical application.
AI tools within a company need to be understood as collectively lead technology, not as an ideology that ignores different points of view.
How to Lead for Inclusive AI Implementation
Because implementing AI involves people as well as machines, it’s important for leaders to maintain a balance—alternate between revenue-boosting projects and more exploratory, fun ones. Coders need motivation, too.
A company culture that embraces AI is also focused on cross-mentoring, not just top-down mentoring. It’s important to spread new ideas and various expertise, as well as experience. Coders in charge of information flow must have opportunities to talk to businesspeople taking care of profit flow.
One way to do this is through frequent, short check-in meetings for AI insight. Stakeholders can do their main work in the environment that suits them best, but they still have plenty of opportunities to communicate with the other side.
AI Relevancy
Anytime your team starts a project, it should be with a clearly defined problem. Only then can you reverse engineer a relevant solution.
When it comes to AI business ideas, engineers often get distracted by accuracy over relevancy—yet another reason there should be communication with the business side, which tends to tip the scales the other way. Failing to consider relevancy (real customer needs and solutions) can be a major threat to AI. There are endless examples of this, such as Google Health, which thrived in a simulator, but failed in real life.
With all the AI hype, many customers want digital solutions—and accurate predictions—fast, but aligning strategy and technology is tricky. The challenge for leaders is to keep things problem-centric, not solution-centric.
How to Keep AI Business-Related and Relevant
AI models need both relevancy and accuracy. They should also depend less on theoretical models and more on problem-centric features, like image-based data.
In fact, we’ve found that AI-focused companies dealing with image-based products like fashion and cosmetics have it easier than manufacturers who prioritize fast and accurate AI models. The reason? The latter often fail to address value for the client—in other words, relevancy.
The way to stay relevant is to start with the end in mind. Use AI only when the problem is clear and a solution offers value.
Next Article
A New Threat to the Digital Age, the Functionally Illiterate Can’t See the Whole Picture
9 Technology Myths: Reprogramming Business Leaders for the Future
AI Insight from Empathy
As smart as machines are, only humans can provide cultural training. And that training is how machines learn to interact with humans in a specific context—in other words, how we achieve artificial empathy.
Consider autonomous vehicles interacting with pedestrians. AI-powered driving systems are taught how to classify pedestrians demographically based on features like shape and clothing color. While this information may seem irrelevant, consider this: Different cultures observe different behaviors. Assertiveness and politeness manifest in road interactions, such as how Italian vs. Japanese pedestrians cross the street (boldly vs. timidly).
Automated vehicles need empathetic input from humans to understand that different pedestrians may interact with them—and the road—differently.
While AI is great at pattern recognition, humans have a unique ability to spot subtle details and make emotional interpretations of what we observe.
Training AI for Empathy
Empathy within AI or by AI opens new business opportunities, as it did for Bombfell, an AI-based software that chooses clothes for people (mainly men).
Businesses hoping to use AI should have a realistic understanding of what AI is and does. But it’s equally important for coders to understand the big picture of business. Leaders need to identify those gaps within a team and help balance differences with empathy themselves.
AI use in organizations will only accelerate going forward. To manage that change, the ultimate AI insight for leaders is the importance of real human empathy and creativity—as well as inclusion and relevancy. Only then can they enhance the value of AI teams to both enhance their product or service and achieve greater corporate goals.