Now that The Active Power of Touch has closed, I am revisiting feedback while preparing for an upcoming conference presentation. The connections visitors made with the artworks were deeply personal and tangible.

Together, artists and visitors discovered that touch is a powerful sense that encourages physical interaction in a shared space.

Physical interaction seems important now as a balance to our digital world. This post reflects how I am thinking about physical touch, and digital, AI as modes with complementary uses. 

Exploring the Differences Between Digital Intelligence and Human Sensory Experience

AI is digital, and sensory inputs like touch are analog. Both are different ways of knowing that offer humans distinct perceptions of the world.

As a blind person, I write about making observations with input from the four nonvisual senses: touch, hearing, smell, and taste. I also write about using AI to obtain visual information by prompting it to describe digital images found online.

Sometimes, I use AI to retrieve information, but it remains digital. Unlike AI, the human senses process is input from real life even when humans aren’t interacting with touch screens or smart speakers.

In 2026, AI systems can produce visual content such as images and videos, generating audio like spoken words or music, and creating written text. Touch is the only sensory output that AI models cannot produce.

There are technologies capable of delivering tactile experiences. For example, a refreshable Braille display connected to a computer or smartphone converts on-screen text into Braille that can be read by the fingers. Smartphones have built-in sensors that provide haptic feedback when receiving notifications, and 3D printers can turn digital designs into tangible objects.

I have not found AI models that can produce tactile output like Braille, or haptic vibrations. While AI cannot manufacture physical objects, I have read about people using vibe coding to access 3D modeling software. However, these 3D digital models are not tangible replicas. If you add a final command to the prompt, a 3D printer could then transform digital instructions into a physical object by melting and shaping plastic. AI can tell me about touch, but it cannot produce an object that I can touch.

Even though AI is limited to exploring digital information, which is incredibly useful. When I want to retrieve information from a website or social media post, I may use AI alongside assistive technologies like a screen reader (voice output), and a refreshable Braille display (tactile output).

When I asked Google Gemini to find scholarly references about tactile perception, it listed key papers that were frequently cited in experimental psychology. Gibson (1962) described touch as seeking information through the movement of the fingers and the hand; this movement is combined with receptors in our skin that sense temperature and pressure. Both the joints of the hand and the receptors in our skin provide rich tactile sensory input that is processed in our brains.

Lederman & Klatzky (1987) recorded and analyzed the ways that people touched objects. They described specific hand movements to explore objects.

For instance, “Contour Following” means using the fingers to trace along the edges of an object to identify its shape. Hand movements differ when people encounter contrasting textures. People stroke or squeeze soft objects, but they tap hard objects.

I leveraged the strengths of AI in finding and summarizing this information about how humans use the sense of touch. Then I searched for these citations and read the key works for myself. AI is valuable for searching databases of scholarly literature, and it can synthesize large amounts of information.

I was inspired to write this post after reading about a professor who asked an AI bot to suggest a class assignment that cannot be completed using AI (Carter 2025). The response to her prompt was to assign a museum visit and have students physically interact with the objects on display. While AI can process information, it cannot replicate the embodied experience and original inquiry that occurs when a student interacts with a physical object.

Professor Carter, and other scholars in museum studies have explained how people form connections when they look at objects and handle them.

Comparing AI to the input of human senses is like comparing apples to oranges. They are two entirely separate ways of knowing — each with their own advantages and disadvantages.

AI can synthesize information and extract key facts that answer a question. The citations of literature in experimental psychology were found to answer my question about tactile perception. Yet, AI cannot produce tactile output. While I’m familiar with AI’s ability to generate audio, I haven’t encountered AI models that produce tactile output beyond providing instructions for a 3D printer.

The human senses give us a rich experience of the world, but they cannot quickly synthesize digital information. As a blind person, I naturally rely on nonvisual senses like touch and hearing to make observations in the world around me.

Through visitor feedback from The Active Power of Touch, I have learned that many people connect to the world through touch. Tactile experience cannot be recreated digitally, and visitors got excited when they were encouraged to touch art.

Now, I know that touch equals engagement. I have heard museum people talk about trying to increase engagement. 

Reframing touch as a benefit for everyone can help museums to compete in a digital AI-saturated world.

References

Carter, S. A. (2025, November 18) Teaching with Museum Collections in the Age of AI American Alliance of Museums.

Gibson, J. J. (1962). Observations on active touch. Psychological Review, 69(6), 477–491

Lederman, S. J., & Klatzky, R. L. (1987). Hand movements: A window into haptic object recognition. Cognitive psychology, 19(3), 342–368.