The Predictability of Human Behavior and the Challenge of AI Workflows

The Predictability of Human Behavior and the Challenge of AI Workflows

Human behavior has always been a subject of fascination and study. Understanding why people act the way they do and predicting their future actions has long been a goal for researchers across various disciplines. Recently, a group of network scientists from Northeastern University made an intriguing discovery - human behavior is 93 percent predictable.

Led by Distinguished Professor of Physics Albert-László Barabási, the team analyzed the mobility patterns of anonymous cell-phone users. Their findings revealed that despite the significant differences in travel patterns, most individuals are equally predictable. The predictability refers to the probability of foreseeing an individual's future whereabouts based on their previous trajectory.

Interestingly, this predictability holds true for both those who frequently travel long distances and those who prefer to stay close to home. Furthermore, the regularity and predictability of individual movement do not differ significantly across demographic categories such as age, gender, language groups, population density, and urban versus rural locations. It seems that regardless of time and distance, people follow a simple pattern and have a strong tendency to return to locations they have visited before.

This research sheds light on the underlying patterns that govern human behavior and challenges the notion of spontaneous decision-making. While it may seem obvious that people exhibit predictable behavior, the study provides empirical evidence to support this claim.

On a different note, the field of artificial intelligence (AI) has been making significant strides in recent years. However, one major challenge faced by AI workflows is the "hurry up and wait" phenomenon. AI advancements often only affect certain parts of workflows, leaving the rest of the process unaffected and causing delays.

In many large companies, there are layers of approval and sign off, which can involve legal considerations. Even if generative AI is used to speed up content creation, the content may still sit idle for days, waiting for approval. Recognizing this issue, some creation platforms have incorporated functionality for machine approvals. However, it is more efficient to have a centralized and independent review process.

To address the "hurry up and wait" problem, it is necessary to focus on the slowest step in a workflow. By using AI to make the slowest step much faster, the overall workflow can be accelerated, even if subsequent steps still rely on human involvement. This approach allows for incremental improvements and adds value to the entire process.

However, implementing AI solutions across multiple product categories requires substantial capital and early competition. It is a full-stack approach that requires companies to invest in various aspects of the workflow to ensure a seamless integration of AI technology.

In conclusion, the predictability of human behavior and the challenges of AI workflows present interesting insights into two different realms. While human behavior may seem unpredictable on the surface, research shows that there are underlying patterns that govern our actions. On the other hand, the "hurry up and wait" problem in AI workflows highlights the need for comprehensive solutions that address the slowest steps in a process. By focusing on these areas and incorporating AI technology, companies can optimize their workflows and achieve greater efficiency.

Three actionable pieces of advice stemming from these insights are:

1. Take advantage of the predictability of human behavior in various fields, such as marketing, transportation, and urban planning, to improve decision-making and resource allocation.
2. Identify the slowest step in your workflow and explore how AI technology can be leveraged to accelerate that particular stage, thereby improving the overall efficiency of the process.
3. Consider investing in a full-stack AI solution that covers multiple aspects of your workflow, allowing for seamless integration and avoiding the "hurry up and wait" phenomenon.

By combining the knowledge gained from understanding human behavior and the advancements in AI technology, organizations can unlock new possibilities and drive innovation in various industries.