IA’s role in AI
I live in San Francisco with two Google Homes, two Amazon Dots, one Amazon Echo, and six housemates who adore artificial intelligence applications. My boyfriend, an enthusiastic Google Home advocate, insisted on placing one in our room so he could configure it with smart lights.
“We really don’t need this,” I told him.
“You’ll like it. Trust me,” he said.
A few days later, I was having Google Assistant report the day’s weather and increase the lamp brightness to 90%. One door down, a housemate asked for an overview of his calendar. In the kitchen, another housemate made a song request: “Alexa, play ‘Feel It Still’ by Portugal the Man.” (That song is unreasonably catchy.)
In its current stage of iteration, products like Google Home and Amazon Echo are far from being a “necessity.” If someone were to strip me of my Google Home, I wouldn’t feel much of a loss. But this may one day change. Intelligent digital assistants are growing more sophisticated and powerful, and AI technology as a whole continues to improve and evolve.
Artificial intelligence (AI): computer systems able to perform tasks that normally require human intelligence, such as visual perception, speech recognition, decision-making, and translation between languages.
Where does IA come in?
AI applications require a clear set of use cases and a solid information architecture. Information architecture, hence, plays a crucial role in making AI more powerful.
Information architecture (IA): IA focuses on organizing, structuring, and labeling content in an effective and sustainable way. The goal is to help users find information and complete tasks. To do this, you need to understand how the pieces fit together to create the larger picture and how items relate to each other within the system.
AI allows computers to tackle the kinds of problems that typically fall to human cognition. Since every AI program interfaces with information, the better that information is structured, the more effective the program is.
Seth Earley, founder and CEO of Earley Information Science, explains that for AI programs, “ontology” is the domain of knowledge and IA mechanisms for accessing and retrieving answers in specific contexts.
Earley points out that the ontology must include “common sense” knowledge concerning real-world logic and relationships. These relationships, facts, and terms are the points of inference that help AI programs answer problems. This will help make the AI more forgiving and the user experience more smooth. For instance, if a user makes a request with a slight wording variation, the AI system will still be able to “reason” through this and understand.
As the AI hype train grows, we see organizations of all types using AI to improve their digital means of engaging with users. An example area of focus would be on customer facing information systems such as search and content personalization methods. Having an experience that is customized to a user’s particular interests, needs, and concerns increases the value of that experience. Getting the recommendations right means leveraging a user’s context, characteristics, and place in the search journey. This will require the incorporation of multiple data sources, ontologies, and reasoning algorithms.
Behind a seemingly simple tool is actually layers upon layers of information, organized by careful human effort and thought. The work of an information architect is imperative for the success of AI.
Product designer Timothy Jaeger sums it up well:
“In short, I view the challenge for our discipline as ‘the architecture of understanding.’ We are still responsible for helping our users to understand where they are, what they’ve found, what to expect, and what’s around. But we can’t afford to focus solely on users. We must also use our skills and knowledge to help our clients and colleagues understand what’s possible and desirable. Our expertise with categories and connections will come in handy, but we must also dig deep into culture and cognition.”