At this year’s AI Summit – San Francisco, the buzz of conversations from people eager to learn and share was palpable. The two-day event showcased data and AI vendors as well as brands that are making significant headway in the utilization (and creation) of AI products. Just meeting the vast array of companies from various industries at this event underscores how many businesses are realizing the potential of AI as the technology matures. Inspired by all the learnings, I wanted to share my key takeaways:
1. AI is a cross-industry capability
Many businesses across industries (CPG, media/entertainment, fintech, healthcare, etc.) are investing in discovering how AI can improve themselves. Industries are realizing that the power of AI exceeds the capability of just robotics process automation (RPA). In an era where consumers judge based on their last best experience regardless of where it came from, using AI will become a requirement for providing service and personalization.
2. Experience design for AI products is an opportunity for all
The current AI product landscape has a focus on the technical aspects, and thereby loses sight of the usability of the product. In that regard, AI is progressing like many other technology-driven innovations: the idea comes first, the technology comes second, and the design/usability follows after the conception of product differentiation evolves, from “what can the technology do?” to “how easy is the product to use?” With AI, businesses are still focused on products. Tools such as design sprints can help them incorporate usability along with product design rather than treating usability as a later-stage priority.
3. The AI value chain needs a strategy
AI is more than just a technology; it’s a value chain that can affect every aspect of how a business operates, including how it delivers value with its partners and customers. For businesses, the magnitude of the AI value chain poses a challenge. They know that various functions inside their organizations can be changed by AI. But they don’t yet know which parts of the AI value chain align with each AI-enabled function.
4. AI Talent is a drawback for many
Most companies feel a shortage of skills within the AI field as they try to build out a data, insights, and/or AI team. The most commonly sought roles are data scientists and data engineers. However, businesses need to better understand if they truly need those skillsets because the platforms being introduced require less technical expertise to set up and garner insights. One way businesses can determine the level of in-house data/AI knowledge that’s needed is to evaluate whether the business is using AI for ancillary services or building AI products.
5. AI is iterative and complementary
Machines can learn by themselves (to a degree) – but “machine learning” does not mean machines become highly intelligent and self-aware as science fiction movies would suggest. Here is a simple analogy: AI is like a child growing up – they learn from those more experienced like a child learns from mom or dad and experiments with their surroundings. Machine learning models are the foundation of AI and need to be trained with rules and different data sets (real and synthetic). In time, those machine learning models will mature to start “thinking” alongside humans to a greater and increasingly reliable level.
6. A flexible AI platform is needed
A flexible platform is essential to enable data scientists and data engineers to work efficiently, especially when it comes to managing an entire data pipeline and extracting value from it. As my colleague Yingwu Gao, Vice President of Product Engineering and AI Practice, said, “In the game of big data, whoever effectively gains insight from their data will win. But they first need a reliable way to turn that data into an asset. That’s where AI comes in.” Check out more on this topic from Yingwu here.
What Does This All Mean – AI is Rapidly Evolving
As digitally-led companies establish AI platforms, we’re going to see a plethora of new, applied use cases emerge in the coming years. Today’s core use cases like fraud detection or automated ticketing just scratch the surface of AI’s potential. With the benefit of new use cases, businesses are (finally) going to be able to tap into the wealth of insights currently sitting in data lakes.
For consumers to realize the potential of AI, businesses should apply it to utilitarian uses that benefit humans – such as increasing productivity, heightening personalization, or simplifying decision points. When businesses figure out which (real-world) problem to solve, they can then identify the right data to operationalize to give them the insights needed to tackle the problem. By investing in AI, businesses will be able to capitalize on AI’s unmatched speed, deeper reach, and broader scale to help people accomplish not only the tasks that can be done easily, but also the ones that can’t.
Last, I believe that solving real problems with AI is best accomplished by creating lovable experiences for people. How? By designing solutions with people at the center. This imperative is especially acute for AI, a technology that is not often associated with being lovable and definitely not human-centric even though it can be. In my opinion, this is wholly true regardless of whether you are just getting started on your AI journey or are trying to extract more value from your AI platform. The products we build should provide lovable experiences and outcomes for those they serve, and the more seamlessly they can fit into our users’ lifestyles, the better job we’ve done – AI has the power to intelligently and accurately inform how to accomplish this, so why wouldn’t we tap into it?