What’s Missing in Artificial Intelligence?
Category:UncategorizedEveryone is talking about artificial intelligence, but we’re missing some key changes that it will unleash. As an optimist, I believe it will be a catalyst for changes that will help all of us to learn faster and achieve more of our potential.
Focus on the data
Where do I start? Let me start by noting that, in all the conversations about artificial intelligence, very few people are talking about the data. Most people don’t recognize that AI is actually extremely stupid without data. Data is the fuel that shapes the intelligence of AI. Everyone seems to assume that more and more data will be available as AI evolves. But is that assumption valid?
The erosion of trust
Here’s the challenge. Trust in large institutions is rapidly eroding around the world. As trust erodes, people are going to be less and less willing to share data about themselves and their activities with large entities. People will increasingly embrace technology tools that can help them be much more selective in providing access to their data. This continues to be a big opportunity for a new kind of business that I called “infomediaries” – businesses that will become trusted advisors and manage our data on our behalf (I wrote about this in my book, Net Worth).
Of course, there will still be a lot of historical data on the Internet accessible to AI but, in a rapidly changing world, the most valuable and useful form of data will be data about current activities and preferences. That data will likely be harder and harder to access when people are less and less trusting of large institutions.
What about tacit knowledge?
There’s another issue regarding data. In a rapidly changing world, data captures a shrinking portion of our knowledge. There’s an important distinction that needs to be made when we seek to understand knowledge. There’s explicit knowledge and tacit knowledge –a distinction first made by Michael Polanyi back in the 1960’s.
Explicit knowledge is knowledge that we can express and communicate in words. Tacit knowledge is knowledge that is embodied in our actions, but that we would find very challenging to express. It’s about knowledge that we acquire when dealing with real-life situations and seeking to find ways to have increasing impact. Some tacit knowledge is long-lasting – it involves mastering enduring skills and practices and cannot be acquired by reading books or listening to lectures. Those who have mastered these skills and practices find it very hard to explain everything they do.
Here’s the challenge – in a rapidly changing world, tacit knowledge increases in proportion to explicit knowledge. We are increasingly confronting rapidly evolving situations and developing practices that will help us to have more impact in these situations, but the knowledge we are developing is largely tacit knowledge embodied in practice. We have a hard time expressing in words and numbers what we have learned.
AI is very good at capturing and studying explicit knowledge, but tacit knowledge is largely invisible to AI. Yet, more and more of our new knowledge is tacit knowledge. If we are serious about learning faster, we will need to find ways to connect with people who have developed new tacit knowledge and build deep, trust-based relationships with them so that we can closely observe their practices and gain insight into the tacit knowledge that is shaping their practices.
So, back to the topic of data. Data is the fuel that powers AI, but data is generated through explicit knowledge. If explicit knowledge is a shrinking portion of our total knowledge, the data fueling AI will be a smaller and smaller portion of the knowledge that is rapidly evolving in our world.
The trends towards decentralization
To learn faster in a rapidly changing world, we will need to build deeper trust-based relationships and to provide improved access to tacit knowledge. This is why I believe we will see an increasing trend towards decentralization of our economy and society. I have written about fragmentation and concentration trends in our economy here.
Decentralization will provide the context for enhancing the potential of AI. It will help us to build the trust that will motivate us to share more of our data. It will also help us to gain more access to the tacit knowledge that can expand the value of our data. The AI apps that will ultimately add the most value are those that focus on gathering access to richer, real-time data about the specific contexts that are most important to the users.
But decentralization stands in sharp contrast to the current direction of AI apps, which are aggressively seeking to gather more and more data, wherever it resides. These apps are providing the insights from that data to a broad range of potential users, regardless of their situation or motivation.
The rise of the Contextual Age
We are in the early stages of a Big Shift from the Industrial Age to the Contextual Age, as I have written here. In short, we are moving from a scalable efficiency model where the key to success was offering highly standardized mass market products and services to a scalable learning model where the key to success is understanding the rapidly evolving contexts of individual customers and organizations, and then offering rapidly evolving tailored products and services to meet their individual needs.
In this Contextual Age, our need for data shifts. Rather than seeking to gather as much data as possible on a global scale, we need to become more focused on gathering richer, real-time data about the specific contexts that matter the most to us.
This will be a significant shift in the direction of AI. Those AI app developers who understand and pursue this shift will be the ones to create the most value. They will also unleash a virtuous cycle because the users of their apps will see significant value in terms of insight that matters to them when they provide their data. This will deepen the trust of AI users and motivate them to provide even more data in their quest for even more value.
Bottom line
AI has significant potential, but only if we recognize the growing challenge of accessing data and the tacit knowledge that is one of the key results of accelerated learning. AI can help us in addressing these challenges but only if we expand our focus to explore what data is most meaningful in a rapidly changing world. As we begin to see the data that matters the most, we can then focus on how to access that data in ways that will deliver increasing value for all and increase our access to even more data that matters.