AI Political Workshop

Grounded in critical tech theory, this hands-on workshop reframes AI from a tool of corporate extraction to a site of collective political action. Participants will move beyond ethical critique to learn technical intervention methods, directly manipulating open-source image and text models to challenge dominant narratives.

Key Learning Modules:

• Deconstructing Systemic Bias: Auditing standard diffusion and text models to visually map out embedded colonial, racialized, and gendered assumptions.
• Critical Data Injection: Actively gathering, tagging, and structuring alternative datasets to counter missing histories or skewed representations.
• Hands-on LoRA Training: Learning the mechanics of Low-Rank Adaptation (LoRA) to train lightweight, custom weights that alter a model’s default behaviors.
• The AI Commons: Exploring alternative governance frameworks, data cooperatives, and decentralized models that put algorithmic control back in the hands of communities.

TimeSession Title & Details

1. Welcome & Setup [ 09:00 AM – 09:30 AM ]

• Onboarding, validation of API keys/LLM platform access, and environment verification.

• Establishment of fundamental data privacy protocols and safe local operational workflows.

2. Prompting for Policy Foresight [09:30 AM – 10:30 AM ]

• Introduction to structured engineering frameworks (Persona, Context, Task, Format) specialized for public data processing.

• Hands-on lab focused on ingesting complex historical regulatory updates to automatically extract emerging industry trends.

☕ Morning Coffee Break [10:30 AM – 10:45 AM]

• Informal networking and peer-to-peer concept reviews.

3. Argumentation & Logic Stress-Testing -[10:45 AM – 11:45 AM]

• Operationalizing AI agents to simulate hostile or conflicting adversarial perspectives.

• Practical sprint loading draft positions into an evaluation container to run interactive multi-stakeholder simulations.

4. Audience Translation & Localization - [11:45 AM – 12:45 PM ]

• Mechanics behind multi-tier content summarization, structural extraction, and targeted tone adjusting.

• Sandbox lab breaking a dense legislative framework down into technical briefs, media updates, and localized community materials.

🍽️ Networking Lunch -[12:45 PM – 02:00 PM]

• Mid-day operational break and open technical discussion.

5. The Guardrails & Ethics Audit - [02:00 PM – 03:00 PM]

• Strategies for identifying statistical hallucinations, structural algorithmic bias, and missing context anomalies.

• Evaluation challenge where participants manually parse automated responses to identify, flag, and patch factual errors.

6. Capstone: The Policy Sprint [03:00 PM – 03:45 PM]

• Live group simulation applying all architectural prompting methods to an unfolding, time-sensitive ecosystem crisis.

• Collaborative drafting, rigorous stress-testing against synthetic agents, and rapid generation of multi-channel public communications.

7. Debrief & Toolkits [03:45 PM – 04:00 PM]

• Final analysis of capstone outputs, structural review of logic pipelines, and distribution of baseline production-ready system prompts.