Four Ways to Build an AI-First Culture

The AI of today and tomorrow will look very different from the AI from yesteryear. To embrace its limitless opportunities, businesses need to adapt an AI-first culture. Here's how.

By Molly Golden, emerging technologies messaging lead for artificial intelligence, Dell Technologies

To The Future

It’s 2030. Laura is an artificial-intelligence (AI) native, living in Buenos Aires. She grew up with an AI assistant and studied AI design and programming. Recently she’s been mentoring her Generation X colleagues, helping them partner more closely with AI. She doesn’t know when they’ll stop writing emails themselves, but she believes that by learning how to effectively judge what machines can and cannot do, her colleagues will come closer to realizing a new world of work, in which bots will be seen as digital colleagues and humans will focus on high-order tasks.

This is a day-in-the-life scenario, envisioned by the Institute for the Future (IFTF) in its seminal Future of Work report.

To The Past

AI was birthed via repeatable tasks traditionally done by people that were offloaded to machines. Math genius Alan Turing cracked Enigma during WWII by building a computer that could decode patterns far faster than any human. He predicted that further down the line, machines will become so intelligent, they’ll imitate humans and it will be nearly impossible to decipher between who or what is answering the questions. While this gray area is fascinating, it’s also controversial: AI may imitate humans so well that one day they’ll take our jobs. In reality, however, the situation is far more nuanced than that.

To The Present Day

As the technology has advanced, humans have begun to partner with AI. This partnership has the potential to advance the world of work at a systemic level, if it’s accompanied by a substantive shift in mindset and culture. While the necessary computer power has already arrived, along with its capacity to handle exponential varieties and volumes of data at velocity, businesses will need to do their part to redesign the work culture and harness the collaboration between people and machines.

Here are four ways to do this.

1. Engendering Trust in the Technology

A recent survey of AI adopters from O’Reilly Media showed that the greatest impediment to successful AI adoption isn’t the technology, it’s an unsupportive culture. So, how can organizations win support? It starts with earning institutional trust in the promise of the technology, in terms of how the technology is used and toward what end. If the path forward is AI as our digital co-workers—built with Third-Wave digital assistants that encompass our human perception system—we’ll need reassurance that we’re not working with “black boxes,” or systems where the inputs and outputs are understood, but the internal workings are a mystery to the users and, frequently, the creators.

To inspire trust, transparency is key. We must know how AI systems are built, as well as how they store, handle, and use the data they collect. To trust a decision made by an algorithm, users need to know these algorithms are reliable and contextually aware (i.e. take into account society’s values and norms), the sequence of events that lead to the decisions can be challenged, and controls are in place to ensure no harm can be caused. AI decision-making shouldn’t be able to hide behind inscrutable code.

2. Adopt a Test-and-Learn Approach

AI isn’t a plug-and-play technology, yielding instant results. It’s complex and requires continuous learning and development.

Companies that want to embrace an AI-oriented culture will need to shed the old mindset of having fully baked ideas and immediate success—and in its place adopt a test-and-learn mindset, in which mistakes and failures are reframed as opportunities for learning and discovery.

Starting small, companies can gain a lot of actionable insights by conducting simple business experiments and applying these insights as they scale in the future—as Dell Technologies knows full well when it established its DevOps practice and drove a shift in its culture, towards a more iterative development approach, during which processes are continally revised, lessons are learnt, and new avenues are pursued.

3. Redefine What It Means to Be “Smart”

Hollywood Sci-Fi movies have perpetuated the fear that AI will take all our jobs, but the reality is that AI will act as a complement to human skills. Today, machines might excel in carrying out repeatable tasks, but it’s highly debatable they’ll develop emotional intelligence any time soon. In the age of AI, imitable human skills, like communication, empathy, compassion, and listening, will be even more valuable—if these people can also comfortably interface with machines, as Laura in Buenos Aires can.

Of course, hard skills will remain important. An organization that is AI-oriented will ensure everyone from leaders down in the organization is AI literate, meaning they understand what AI is, how it works, and what it can do for the company. But the magic will happen at the interplay between machines and humans. According to IFTF, “Just as humans need to improve our ability to offload tasks that are better suited for machines, so AI systems will need to be designed to recognize computational limitations and know when to lean on their human partners for help with completing a task.”

4. Write the Job Descriptions of Tomorrow

The proverb “it takes a village” is fitting for AI. An AI-oriented culture recognizes that AI can’t be successful in a silo. It requires a multi-faceted team with a mix of skills and perspectives. This team includes the C-suite, data scientists, data analysts, AI trainers, AI explainers, among others, all working towards a common goal.

Interestingly, some of the participants that attended an IFTF expert workshop in March of 2017 imagined a future in which 85 percent of jobs in 2030 haven’t been invented yet. This means workers will need to get accustomed to learning on the fly now and adapting their skill sets in novel ways.

Realizing AI’s Limitless Possibilities

The AI of today and tomorrow will look very different from the AI from yesteryear. We’ve moved beyond First Wave AI and we’re now powering through Second Wave AI, with our eyes fixed on Third Wave AI and its limitless opportunities. As we ride the crest of these waves, an AI-first culture will prove to be the linchpin of a future powered by intelligent technology.