The Final AI Tensions: Personalization, Objectivity and Power
FM (Friday Morning) Reflection #28
In this series
This is the fourth in a series of FM (Friday Morning) Reflections that explore the tensions created by generative AI in media, content creation, and the consumer-brand relationship. As AI tools become more integrated into daily life and work, AI literacy is no longer optional—it’s essential. But AI literacy goes beyond knowing how to use AI tools. It requires grappling with the deeper ethical, creative, and social implications that AI raises.
So far, we’ve examined AI’s impact on authenticity, ownership, corporate innovation, and public responsibility. Today, we’ll explore the final three critical tensions (bringing the total to twelve!) that will shape how we interact with AI in the years to come.
Personalization vs. Manipulation
Personalization enhances user experiences by tailoring content based on our preferences and behaviors. Each time we trade information for convenience—whether knowingly or unknowingly—we must ask: Is this a fair value exchange?
The digital content ecosystem thrives on attention as its primary currency. AI excels at recognizing patterns in unstructured data. Used responsibly, AI-driven personalization systems can enrich experiences, streamline workflows, and foster engagement.
These systems improve when they have access to more information about us, and this drives the quest for more instrumentation and data about our lives. However, at a certain point, the line between helpful personalization and surveillance becomes increasingly difficult to distinguish.
These same mechanisms can also be used to manipulate users, subtly shaping opinions and behaviors. Take the case of DeepSeek, a new AI model and application criticized for censoring answers that might conflict with topics sensitive to the Chinese government. While the model’s capabilities are technically impressive, such biases highlight the dangers of AI systems subtly shaping discourse without users realizing it. With AI-powered assistants playing an increasing role in decision-making, the risk of manipulation is growing.
Here’s the question to always keep in mind: When we use AI tools, to what degree are we making independent choices based on objective facts, and in what ways do algorithms subtly guide us toward certain outcomes? AI literacy requires understanding these dynamics to navigate this evolving landscape wisely.
To add to this tension, privacy lapses with DeepSeek have reportedly lead to the exposure of deeply sensitive data including user prompts and personal information. Whether by accident or by intention, these incidents raise urgent concerns about security and the hidden actors who may exploit this data.
Objectivity vs. People-Pleasing
At their core, large language models (LLMs) are predictive systems. They generate responses based on probabilities influenced by training data, optimization goals, and user feedback loops.
Many AI models are designed to be "helpful" and "friendly" and our feedback to them guides their output. This is both a feature and a bug, because this tendency becomes problematic when AI avoids uncomfortable truths or reinforces user biases to maintain engagement.
If a model’s main directive is to provide agreeable answers, will it sidestep hard facts? Will it tell users what they want to hear rather than what they need to know?
This is a fundamental tension. AI can be a powerful tool for expanding knowledge and critical thought, but it can also act as an echo chamber, reinforcing selective perspectives, all while delivering answers with enthusiasm and confidence.
Democratization vs. Inequality
Today, broadband is considered critical infrastructure, and we recognize access as essential in boosting economic and societal outcomes. Tomorrow, AI will be critical infrastructure too.
AI has the potential to democratize access to technology by enabling people to interact with advanced systems through natural language and voice commands. In FM (Friday Morning) Reflection #10: Getting from Where We’ve Been to Where We’re Going, I explored how AI offers a “superpower” that expands opportunities and empowers users who previously lacked access to technological tools.
However, AI also risks exacerbating existing inequalities.
If access to advanced AI models, high-quality datasets, and computational resources is limited to those with power, a divide between those equipped with AI tools and literacy and those will grow, creating new disparities in economic and social mobility.
Ensuring that AI benefits the largest number of people requires proactive efforts—and integrating AI literacy into educational curricula, workforce training programs, and public policy initiatives is a good start.
Where Do We Go From Here?
Over the course of my last four Reflections, we have explored twelve fundamental tensions shaping our relationship with AI.
Identifying these tensions is just the first step. Now, the real challenge begins: developing an actionable AI Literacy Framework, because balancing the issues raised here doesn’t rest solely on AI developers—it also requires users to understand how the technology works, critically evaluate AI-generated information, and to engage responsibly.
This won’t be a static, one-size-fits-all solution. AI literacy is an evolving practice—one that requires iteration, adaptation, and collective insight. I’ll continue refining this framework and will share my findings as they take shape.
What are your thoughts? How do you see these tensions playing out in your own work and life? Let’s continue this conversation with your comments and feedback to this post.
Have a great weekend!
Photo Bonus
My recent Reflections on the tensions inherent with AI featured black and white photos of scenes from Cherry Street Pier in Philadelphia. In several, the architectural elements of the structure frame the scene, much as I hope we can use an AI Literacy Framework to help provide context for how to use AI. Now we zoom out to to the larger view in which life was taking place inside the pier building as I walked through. I was delighted to find a quartet playing at the entrance and took a minute to enjoy the scene, surroundings, and sounds. The man greeting visitors to the pier, seated at left, also seemed to appreciate the addition of music to the space as he looked on.