FUTURE_TOKENS

Future Tokens are named, composable instruments that target specific blind spots in AI output—gaps that coherence optimization typically suppresses during token-level generation. Each operation targets a distinct blind spot type: opposition gaps (antithesize), hidden assumptions (excavate), missing dimensions (dimensionalize), conspicuous absences (negspace), rhetorical fulcrums (rhetoricize), cross-example patterns (inductify), structural echoes (metaphorize), actionable handles (handlize), and position integration (synthesize). Operations teams use these skills to build AI-assisted thinking workflows that surface information the generation process systematically omits. Compatible with Claude Skills, Cursor, and Codex workflows, the skillset installs as reusable cognitive operations available as /commands.

[{"step":"Define your role and objective. Replace [ROLE] and [OBJECTIVE] in the prompt template with your specific context (e.g., 'Act as a UX researcher analyzing friction points in [PRODUCT] onboarding').","tip":"Be precise: instead of 'analyze friction points,' specify 'identify the top 3 drop-off points in the first 30 seconds of onboarding.'"},{"step":"Choose the FUTURE_TOKENS components to prioritize. Focus on one or two (e.g., 'antithesize' and 'synthesize') if the topic is complex, or use all four for a holistic view.","tip":"For controversial topics, lean into 'antithesis' to challenge assumptions. For emerging trends, 'metaphor' can help crystallize abstract ideas."},{"step":"Run the prompt in Claude Skills or a similar AI tool. Use the 'Skills' feature to save this as a reusable template for future iterations.","tip":"Iterate: After the first output, ask the AI to 'refine the synthesis section to focus on [SPECIFIC OUTCOME],' or 'add a 5th step to address [GAP].'"},{"step":"Extract actionable insights. Highlight the 'Synthesis' section and use it to draft a memo, roadmap, or decision document. Share with stakeholders for feedback.","tip":"Use the 'Excavation' section to flag risks or opportunities that might otherwise be missed (e.g., regulatory blind spots)."},{"step":"Automate repetitive use cases. For example, if you frequently analyze competitor strategies, create a Claude Skill that auto-fills [ROLE] as 'competitive intelligence analyst' and [OBJECTIVE] as 'antithesize the gaps in [COMPETITOR]'s product positioning.'","tip":"Combine with other skills: Use FUTURE_TOKENS to generate hypotheses, then feed the output into a 'DATA_VALIDATION' skill to test assumptions with real data."}]