Ai, Computation and the Human Touch (Thinking, Sensing, and Making)
Onur Yüce Gün, PhD
Director of Computational Design at New Balance | 100x Lecturer and Keynote Speaker in Computational Design and AI
"Live" thoughts from the PI Apparel STRIDE 2025 closing panel
I sat in the closing panel of the Stride 2025 event of entitled Creating Synergy Between Traditional Craftsmanship and Digital Innovation.
Below are the three key points that flowed from my mind during our discussion. I used my jetlag to transform those thoughts into words.
Highlight 01: Are Computation and Data Pure?
Computation implies converting ideas from tangible reality into numeric entities. In this process, we translate qualities we understand into quantities—numeric values that work only for aspects of reality that can be directly quantified.
But much of meaning and value remains incomputable in that sense.
So while computation suggests a sense of absoluteness and perfection, it inherently involves a kind of reduction. The important aspects of reality and meaning are left out as we use computation to realize designs. Although this might seem like a problem, having this awareness actually helps refine our approach. This realization allows us to use computation in a much more informed way.
Highlight 02: AI and Its Inherent Connection to the Past
AI is inherently rooted in the past because it operates on datasets composed of things that have already happened and have been recorded in digital form. When we train AI, we work with statistically distributed data points, which we often represent using mean distributions or bell curves. Essentially, we are searching within the central region of these curves—meaning we are reinforcing patterns from the past.
But if we’re aiming for creativity, innovation, or novel ideas, our search ideally should happen at the extremes of the bell curve. And if we’re looking for a true leap, it needs to take place outside the curve altogether.
This raises an interesting question: if AI is fundamentally shaped by historical data, can it truly be creative?
That leads to the next point—how we instinctively anthropomorphize AI, which we’ll cover in the third note.
Highlight 03: The Immediate Anthropomorphization of AI
AI has been around for a long time, conceptually dating back to the 1930s, with early implementations emerging in the 1950s. Despite decades of research, mainstream awareness of AI only surged in recent years due to increased accessibility.
As a result, many now equate AI solely with generative AI—a subcategory of the broader artificial intelligence field. This has led to a widespread perception that AI "thinks," "writes," and "creates" in ways similar to humans, which is simply not the case.
At its core, AI is a powerful pattern analyzer and generator. The text and images it produces are fundamentally dictated by statistical patterns rather than human-like cognition.
Large language models use tokens and statistical probabilities to predict and generate sequences of words, while image-generation models, such as diffusion models, rely on noise manipulation and pattern reconstruction to create visuals based on labeled training data.
Both systems operate in ways that are fundamentally different from human thought and creativity, yet the tendency to anthropomorphize AI remains strong.
Main Takeaway: Understanding AI Beyond the Hype
If we step back from the initial awe of AI—whether it’s the speed of generating images and text or the predictive capabilities of synthetic insights—we can develop a clearer, more analytical understanding of these tools.
At their core, AI systems are large-scale machines that process vast amounts of data, but they do not operate autonomously in a meaningful way. The role of interpretation, direction, and insight-building remains firmly in human hands.
While deep learning and neural networks function in ways that might resemble aspects of human cognition, they are not direct replications of the human brain.
AI does not "think" as we do; rather, it operates as a model of the mind, abstracting certain processes but lacking the full biological and experiential complexity of human intelligence.
This is why intuition, experience, and knowledge remain essential. Humans are responsible for curating and refining the instructions given to AI, interpreting its outputs, and using its results in a meaningful and informed way. AI can generate, but we must make sense of it.
Follow me on LinkedIn and YouTube for in-depth understanding of human and artificial intelligence, and their implications for future of life and work.
#AI #DigitalInnovation #Craftsmanship #Computation #DataScience #ArtificialIntelligence #MachineLearning #Innovation #Technology #FutureOfWork #ComputationalDesign #Growth #HumanIntelligence