This experimental prototype was part of my 8-year research journey that led to Workshop Workshop.

Join the current experiment →

MMM

Meaning Making Machine

A transparent multi-tiered robotic device with various white objects, rocks, and small items on its shelves, set against a black background.

The Vision

A detailed black and white illustration of a complex diagram about AI systems. It includes elements like reactions, an AI profile, scaffolds, amulet, god, a miniature labyrinth, training station, body language, scoring rules, historical data repository, text, physical computing, component crowdsourcing, and pre-tâe-belle. Visually, it depicts interconnected sections with handwritten labels, sketches of patterns, people, and various objects, illustrating concepts related to artificial intelligence, data processing, and human interaction.

The MMM represents was the final project for my MA/MSc in Global Innovation Design, embodying my explorations into recommendation models for object customisation. The original vision was to create an environment encouraging participants to engage with various stimuli, uncovering their preferences so the data could feed into a model creating bespoke meaningful objects.

Due to time constraints and budget limitations, I scaled down from a full environment to designing a singular touchpoint: a decision that greatly influenced the project's trajectory.

The Question Behind it

A collage of various geometric and abstract objects, some sketched and some in color, including polyhedra, cubes, cones, spheres, and architectural models, with some icons and technical diagrams interspersed throughout.

The concept stemmed from inquiring about what made certain objects feel meaningful. If I combine two objects with different meanings, does the resulting object become meaningful in a different way? Antiquity is full of meaningful objects, amulets remain popular worldwide even though the beliefs that originally enabled them have changed.

I started visiting museums and interviewing people about their relationships with objects they considered meaningful. Less sophisticated examples like Pandora bracelets were everywhere, but I was envisioning something deeper and more personal.

Technical Implementation

A person interacts with a complex transparent mechanical device with various small sculptures and objects inside, resembling an art installation or experimental machine.

The machine quantified user interactions through a sophisticated system. Participants would choose objects following prompts I provided, then place them on the machine's top tray. Each object had a tag, and sensors beneath the tray identified these tags, capturing user choices.

Creating the combination mechanism required treating it like an architectural project, where each decision influenced others. I chose transparent materials to demystify the process, allowing participants to see the inner workings and enhance their understanding.

Art collage of various black and white objects, including feathers, a bottle, an egg, and a statue, arranged on a white background.

The Modular Amulet System

The MMM became a physical machine designed around creating modular amulets. I used a recommendation engine typically used in e-commerce, building a database of varied objects contributed by artists including Cristina Carbajo, Melisa Leñero, Andriana Nassou, Jessica Gregory, Sadi Yetkili, and Kelly Fung.

To gather initial data, I asked volunteers to classify objects into categories both subjective ("ugly - beautiful") and objective ("small - big"). This approach was crucial for beginning to understand how people ascribe value and attributes to objects.

A collage of objects cut. The body is a round coin, with various tools and parts making up the legs and head.
A collage of various objects including skulls, brushes, and electronic components arranged around a golden ring with green accents on a white background.
Black and white collage art featuring various art tools and a human skull, including paintbrushes, an artist palette, a tube of paint, and a palette knife arranged around the skull.

The Results

A hand holding a mirror reflecting a bunch of decorated stones with various black, white, and blue designs against a clear blue sky.

Creating the final objects posed significant challenges. My initial idea of using elements from various religious deities was set aside due to cultural sensitivity concerns. Instead, I opted for abstract objects crafted from clay using magnets for attachment. While respectful, this choice made it harder for participants to connect with the final object, leading to somewhat underwhelming results.

Colored abstract beads on a transparent plastic tray held in hand outdoors.

What I Learned

A scientist working with electronic equipment and test tubes in a laboratory.

The MMM was more than just a machine, it was an ecosystem uniting various components: object repository, interaction interface, electronics, and combination mechanism. But the technical sophistication couldn't overcome the fundamental issue:

meaningful objects require genuine personal connection, not algorithmic assembly.

Reflecting on the project, I recognise areas for improvement, particularly around understanding users' connections with meaningful objects and ensuring reliable data collection. A key oversight was the lack of feedback loops for continuous model improvement.

Developing MMM in 2015, before concepts like Surveillance Capitalism were widely discussed, I initially viewed behavioural data extraction positively. In hindsight, I recognise the need for more ethical approaches to such practices.

MMM taught me that meaning can't be manufactured through clever algorithms. it emerges from process, relationship, and personal connection, not from the final object itself.

AI and Meaning-Making: A 2017 Experiment's Relevance Today

A detailed diagram and notes about a physical computer system, including a user interface with hand icons, a flowchart of data processing with a Raspberry Pi, sensors, and a Rembie API, and drawings of mechanical components and objects.

MMM was essentially an early experiment in using artificial intelligence to create meaningful experiences - specifically, a recommendation system powered by machine learning algorithms attempting to understand and generate personal significance through physical objects.

This approach now feels remarkably prescient given today's discussions about AI's capabilities and limitations. The system could identify patterns in user preferences and execute sophisticated combinations, but it struggled with something more fundamental: genuine meaning-making requires human processes that resist algorithmic replication.

The project revealed a crucial distinction between pattern recognition and understanding. While the AI could detect correlations in user behavior and preferences, it couldn't replicate the cultural context, personal history, and emotional resonance that actually make objects meaningful to people. The most significant moments happened not when the machine delivered its algorithmic recommendations, but during conversations people had while using it - the human reflection and connection that the technology couldn't capture or reproduce.

This limitation feels especially relevant as we navigate today's AI landscape. Current systems excel at pattern matching and generating outputs based on training data, but they still struggle with the deeper human experiences that MMM attempted to address: genuine understanding, cultural sensitivity, and authentic meaning-making.

The lesson from MMM wasn't that AI is useless for human experiences, but that it works best as a tool for facilitating human connection rather than replacing it - an insight that continues to guide my current work