Bots with an Attitude: Ethical AI Collaboration

2020

A framework for designing intelligent agents that participate as equal community members in collaborative spaces, redefining human-AI relationships through ethical creative intelligence.

Context

Jan Hein Hoogstad was commissioned to develop a vision for online collaboration for Planet B, an imaginative thought experiment centered on remote coordination for a planetary expedition. The challenge was immediate and practical: how do you build a thriving community of people dispersed across the globe who may never meet in person? Rather than treating this as a constraint to overcome, Hoogstad reframed it as an opportunity to reimagine what collaboration could look like when freed from geographic limitations. The project emerged from a philosophical conviction that what matters in collaborative work is effective agency—whether a team member is human or artificial becomes secondary to whether they meaningfully contribute to shared goals.

Problem

Loosely-knitted communities like Planet B face a fundamental challenge: continuous knowledge sharing and coordination across rotating participants without the synchronous presence that traditional teams rely on. Information gets lost, context disappears between interactions, and collective memory remains fragmented. The deeper problem is conceptual—existing approaches treat AI as either a threat to human agency or as passive service tools, missing an opportunity to reimagine human-machine collaboration as genuine partnership. There was no framework for designing intelligent systems that could participate authentically in community spaces while respecting human autonomy and values.

Solution

The solution introduced "bots with an attitude"—intelligent agents designed to participate in chat environments as equal community members alongside humans. Rather than operating behind the scenes, these bots are visible, named participants that perform specific, high-value tasks: acquiring and cleaning data, analyzing information, evaluating quality, and presenting findings back to the community. The bots operate through a state machine architecture where specific functions, called "cassettes," hook into different bot states. This modular design is critical—it lowers barriers to entry by allowing community members to contribute new bot capabilities without deep technical expertise. Each cassette is a reusable, single-purpose function that can be plugged into the system, democratizing bot development itself. The framework redefines artificial intelligence as "ethical creative intelligence"—systems designed to generate relevant insights, combine information meaningfully, and respect community values. An "Organic Governance System" complements the bots by associating contributions to specific agents and adjusting reputation scores based on quality, amplifying valuable contributions while muting low-quality ones. This turns community channels from cacophony into meritocratic discourse.

Outcome

The project fundamentally shifts how we think about collaboration and intelligence. Rather than viewing humanness as the measure of value in teamwork, it establishes agency—the ability to meaningfully contribute—as what actually matters. The cassette architecture proved that bot ecosystems could be genuinely participatory, inviting community members to extend capabilities rather than passively consuming fixed tools. The framework raised essential questions about governance in hybrid human-bot teams: How do we make decisions when both humans and machines are stakeholders? What safeguards ensure bots serve community values rather than narrow optimization? By treating Planet B as a counterfactual space to explore these questions safely, the work provided a blueprint for designing collaborative systems that leverage machine intelligence without surrendering human autonomy. The project demonstrates that ethical AI doesn't mean limiting AI capabilities—it means designing systems where intelligence, whether human or artificial, is accountable to community values.