Technologists may be very adept at creating highly innovative AI projects. But the key is how early they can communicate their innovations to executives who are then tasked with implementing the potential business processes. Getting the most out of AI? It all starts with inspired leadership that understands how it can benefit the customer. As with all technology, it ultimately comes down to the customer.
Addressing customer needs should be part of the package from the start, says Dr. Andrew Ng, founder of DeepLearning.AI and assistant professor at Stanford University. “So check if this idea is even technically feasible, and then talk to prospective customers to make sure there’s a market need,” he explained in a recent webinar.
Ng recounts how “he himself spent a long time working on a project before appointing a CEO. But we found that putting in a leader at the outset greatly reduced the burden of having to transfer knowledge—or having a CEO come in and have to reconfirm what we found. We’ve learned that it’s much more effective to bring the leader out at the very beginning. We then spent three months, six, two-week sprints working with them to build a prototype as well as do extensive customer validation.”
The lesson here is that business leaders and managers need to get involved early in AI efforts to help deliver innovation to the customer. “Current AI development is largely driven by technologists,” notes Reggie Townsend, vice president of data ethics for SAS. “Not only that, but technologists often lead the conversation around AI.”
It’s important that “people from diverse backgrounds take stronger roles in AI, as a whole, to make sure we stay focused on the practical risks and opportunities.” Townsend continues. “We need a diverse set of voices at the table who can think about the macro-level implications of AI beyond the specific applications of the technology.”
Technology companies at the forefront of these efforts are paving the way for collaboration that connects AI development with the customer. “For us, leadership in AI initiatives is shared by technical and non-technical roles,” Artem Krupenev, vice president of strategy at Augury. “Technical roles span data scientists, engineers and AI researchers who are instrumental in developing and refining our AI technologies. Meanwhile, the non-technical design roles of product management and business strategy ensure that technology advancements are aligned with our business goals and market needs.”
There are risks for AI initiatives to stay too long in the technology realm. “Like anyone with expertise or skill in a particular field, technologists will have blind spots,” says Townsend. “They are trained in precision and mathematical accuracy, but applying AI in the real world can quickly go wrong. Being technically accurate does not always lead to ideal outcomes in areas such as equity, welfare and justice. Inaccuracy and bias are ingrained in humanity, and we need experts in the field to sort through the mess to uncover risks and opportunities.”
Augury takes a three-step process to ensure the effectiveness of AI in the enterprise: “normalization, socialization and production,” says Krupenev. “During the normalization phase, we ensure universal access to AI tools and provide clear usage guidelines. In the socialization phase, we share success stories and best practices, fostering a culture of learning and collaboration.”
Finally, he says, “in the production phase, we integrate the use of AI into our standard processes and team structures. This phase also involves continuous improvement as we learn and iterate to improve the impact of AI on our work.”
As AI drives more business decision-making, new roles will evolve, Krupenev predicts. “We have already begun this process by establishing a cross-functional GenAI leadership team. These roles connect technical understanding to business implications and use cases. Non-technical roles such as AI product managers, who take ownership of AI products, and AI business strategists, who integrate AI initiatives into business strategy, will be vital.”
There is also an increase in demand for “hybrid socio-technical roles,” says Townsend. “People with some background in AI development, but also with the knowledge and historical perspective of a sociologist, will be very valuable. Career opportunities in AI are open to more than just technologists, but require a basic understanding of what AI is and isn’t, the practical risks and opportunities.”
Follow me at Twitter.