On June 17, Antonella Varesano, scientific advisor at Asperastra Innovation Lab, presented a paper co-authored with Paolo Gallina, a professor at the University of Trieste, at HWID2026, an international conference on Human-Work Interaction Design hosted by the University of West London. The paper is titled “When Machines Dream at Work: Human-Centered Interaction Design for AI-Augmented Creativity in Industry 5.0.”
The paper’s starting point is simple to state, though less so to resolve: in recent years, Industry 4.0 has made processes more efficient, more automated, and faster. But this push has often pushed an essential question into the background: what role remains for people when systems become increasingly autonomous, opaque, and difficult to truly govern?
Industry 5.0 seeks to revisit this issue by shifting the focus to human well-being, sustainability, responsibility, and people’s capacity for action. The paper by Varesano and Gallina fits perfectly into this context, posing a very concrete question: How do we design an interaction between humans and artificial intelligence that does not reduce the human role to passive supervision, but truly enhances it?
A framework for augmented, non-automatic AI
The paper proposes a model, called the Human-Centered Interaction Framework, built around six principles: human agency and control, transparency and explainability, interaction and embodiment, awareness of the limits of machine-generated meaning, well-being and trust, and human governance.
Put that way, it might seem like a theoretical list. In reality, the point is very practical. These six elements serve to prevent AI and robotics systems from becoming black boxes that people merely use or endure without truly understanding what is happening.
One of the most interesting sections of the paper concerns precisely the issue of meaning. Machines can generate convincing, fluid, even surprising outputs. But this does not mean that they “understand” what they produce. An output generated by an AI system remains open to interpretation until someone interprets it, places it in context, and evaluates its usefulness, meaning, and limitations. This is where a decisive aspect of the relationship between people and intelligent technologies comes into play: in not confusing the ability to generate with the ability to understand.
The Lab as a Testing Ground
In the paper, Asperastra Innovation Lab is not presented as a backdrop, but as a concrete case study. Varesano and Gallina describe Innovation Labs as socio-technical environments in which technologies, practices, and people co-evolve: places where one experiments, prototyped, and observes what happens when a system truly interacts with its users.
This aspect is very important to us. It means recognizing that certain questions about AI cannot be resolved solely in academic papers, conferences, or strategic documents. They are also put to the test in laboratories, workshops, courses, and situations where a technology must cease to be merely a concept and become a concrete experience.
A special role is also assigned to the arts. Not as a mere embellishment of STEM, but as a cognitive tool. Practices such as neural cinema or generative audiovisual systems can make the often opaque behavior of artificial intelligence visible. They compel those who observe and use these systems to do something very human: interpret, contextualize, and decide. Ultimately, this is a logic that is familiar to us in the work we carry out at Lab with local AI: not asking the machine to produce meaning for us, but using it as a tool that opens the door to interpretation, discussion, and choice.
Why it matters even outside the paper
Bringing Asperastra’s work to an international conference like HWID2026 means placing the Lab’s day-to-day experience within a broader conversation about the future of work and creativity enhanced by technology. Not as a curious exception, but as part of a very concrete and timely question: what kind of relationship do we want to build with systems that generate, suggest, optimize, and—increasingly—even seem to “imagine”?
The quality of the future of work will not depend on smarter machines in and of themselves, but on better-designed interactions between people, technologies, and contexts.
It’s a significant shift in focus. It shifts the focus from hype to responsibility, from the machine’s performance to the structure of the relationship.
And that is also why this paper interests us beyond the academic context in which it was presented. Because it addresses work, creativity, design, and autonomy at a time when the greatest risk is not merely misusing these tools, but getting used to using them without sufficiently questioning how they operate and what they demand from us in return.
Read the full paper
