Human-centeredness is at the core of our activities. We believe in this so much that we put it on our website’s header. In short, at Quartz Labs, we believe that AI has the potential to do a lot of good or a lot of harm, that really depends on how we use it. And while putting the genie back in the bottle would be impossible and likely senseless, we want to emphatically steer away from technological determinism: whether we’ll remember the AI revolution as the beginning of something great or something horrible, is really up to us.
But what does this mean in practice? How does a human-centred AI transformation take place, as opposed to one driven by different values?
Let’s start by saying that we are still figuring it out. We are living in a world where things are possible that were completely unimaginable just a couple of years ago. Every day that passes, new technologies and implementation ideas come out. This means that there are no best practices: whatever we do with genAI in organization, we are breaking new ground. Anyone that tells you something different is either being overselling their expertise, or generalizing from just a few experiences. So humility really is the starting point.
There are, however, important guiding principles that we think should guide us in identifying emerging practices.
Start with humility, continue with empathy We mentioned humility in the face of constantly evolving practices. The next step is empathy toward the people that interact with this technology. In any human-centered project, we start by talking to people and identifying their problems. Being problem-centered, rather than solution-driven, ensures that the technology we develop or deploy addresses real needs, not imagined ones. Empathy allows us to understand the context, emotions, and desires of the people who will interact with AI systems. It helps us avoid unintended consequences and design AI that augments human capabilities rather than replacing them.
A sprit of Kaizen The concept of Kaizen, continuous improvement, goes back to Toyota, in a traditional factory setting. The idea is that even the most advanced production system can be continuously improved here and there, and that the people involved in the day to day operation of the system are the ones who best know where improvement is possible. This is why many factories start their day with a kaizen meeting with all line operators. In the context of knowledge-work transformation there are two things we can learn. One, no transformation is truly ever complete. We help our clients deploy the best possible processes, organizational designs and AI augumentations for their innovation teams, but things will always have to be tweaked and adapted as technologies open up new possibilities and new opportunities are discovered. Second, management sets an intention, provides a framework and a business rationale for the transformation. But the ”content” for the transformation, the individual opportunities for improvement, most often come from people involved in day to day operations, and this is why talking to them is so important.
Look for joy and mastery "I want AI to do my laundry and dishes so that I can do art and writing, not for AI to do my art and writing so that I can do my laundry and dishes.” This post went viral a few months ago, and for a good reason. People are scared that AI will take over tasks that they find meaningful and creative, leaving them with only the mundane. A human-centered approach to AI must prioritize the enhancement of human joy and mastery. AI should be designed to automate the tasks that are tedious, dangerous, or monotonous, freeing up humans to engage in activities that bring them satisfaction, creativity, and a sense of purpose. This principle is about recognizing that technology should serve human flourishing, not undermine it. Debbie Lovich at the Henderson Institute has found that employees who enjoy their work are 49% less likely to quit. Can we turn AI augmentation from something people are worried about to a way to bring more joy into people’s work?
Look for marginal labor productivity Daron Acemoglu, the most cited economist in the world, argues that the current trajectory of AI, particularly in automation, could lead to a reduction in the demand for human labor, exacerbating inequality and reducing overall productivity growth. According to Acemoglu, the key to a positive AI transformation is focusing on “marginal labor productivity.” This means developing AI systems that complement and enhance human labor, rather than replacing it. In practice, this involves designing AI tools that assist workers in their tasks, making them more efficient and productive without rendering them obsolete. For example, instead of developing AI that fully automates customer service, we should aim for AI that supports customer service representatives by providing them with better information, insights, and tools, enabling them to offer higher-quality service more efficiently. This approach not only preserves jobs but also enhances them, making work more engaging and rewarding. And we believe that, as things stand as of now (August 2024), human+machine still trumps the machine on its own for the greatest majority of tasks.
What about when the technology will actually be so good that it makes no sense to keep a human doing the task? What when competitive dynamics will make it absolutely untenable NOT to fully automate a task? The answer will depend on the nature of the task and on a given firm’s positioning. Some tasks may eventually disappear, and hopefully they are the ones that won’t be missed. The key activity then will be change management: being able to shift employees’ focus or retrain them based on their skills and passions. For some firms, however, having humans in doing jobs that most companies have automated may turn into a differentiating factor. Most of us are not really able to tell the difference between a real Louis Vuitton bag and a counterfeit, but we accept that the story around the original and the tradition and craft that are behind it are worth the price premium. In the same way, the "human-made" badge may one day command a premium for knowldge-based products and services.
In conclusion, a human-centered, AI-powered transformation is about ensuring that AI serves human needs, enhances human potential, and respects human values. Problems before solutions. Humans before machines. By starting with humility and empathy, embracing continuous improvement, prioritizing joy and mastery, and focusing on enhancing labor productivity, we can steer the AI revolution toward a future that benefits everyone.
Comments