epistemic-libido

name

epistemic-libido

description

Filter, compare, and rank papers, posts, captures, threads, bookmarks, product claims, or research ideas for high-entropy mechanistic insight and underpriced leverage. Use when the user asks for alpha, high entropy, the most intriguing or insightful items, sexy ideas, nerdsnipes, sapiosexual appeal, ideas worth expanding, research-agenda candidates, or an explanation of why an idea is exciting; also use to separate conceptually fertile results from hype, leaderboard gains, derivative work, and product news.

Epistemic Libido

Select ideas that make the mind want to reproduce with them.

Do not rank by importance, novelty, popularity, or benchmark score alone. Rank by whether an idea rearranges the reader's model of the world and gives them a new instrument.

Call the target object a high-generativity mechanistic inversion:

• inversion: violate a reasonable prior;
• mechanistic: expose why the result occurs;
• high-generativity: produce new hypotheses, experiments, combinations, or products.

Establish the Evidence Surface

1. Fix the corpus, time window, and exclusions.
2. Resolve the exact underlying artifact. Do not trust shifted titles, repost summaries, screenshots without context, or social engagement.
3. Deduplicate repeated captures and separate an original claim from replies about it.
4. Prefer the paper, code, dataset, proof object, technical report, or full thread over commentary.
5. State what remains inaccessible or unverified.

When the corpus is large, first remove health checks, operational chatter, duplicates, generic tutorials, and routine launches. Preserve a product item only when it contains a transferable mechanism or a hard implementation receipt.

Normalize Each Candidate

Reduce every item to four fields before ranking:

• Claim: What changed?
• Mechanism: Why does it work?
• Receipt: What would make the claim true or false?
• Implication tree: What becomes possible if it generalizes?

If the mechanism cannot be stated, mark the item as an empirical result or claim rather than inventing one.

Treat novelty as conditional on the user's baseline. Search for adjacent systems, prior work, and known implementations when the answer may demote the idea. If the core thesis is already productized or familiar to the user, preserve only the genuinely new remainder.

Score Heat and Alpha

Score each component from 0 to 3: absent, present, strong, exceptional. Use the numbers to force comparisons, not to manufacture precision.

Epistemic heat

• Surprise: Does it violate a defensible prior rather than merely report progress?
• Compression: Does one principle explain several previously separate observations?
• Mechanistic contact: Does it expose a causal, geometric, algorithmic, or systems-level handle?
• Generativity: Does understanding it immediately suggest further experiments or combinations?

Alpha

• Leverage: Can a small intervention control a large capability or cost surface?
• Transfer: Does the mechanism escape its original benchmark, model, or domain?
• Underpricing: Are the implications larger than the attention they are receiving?
• Actionability: Can the user test, build, or exploit it without waiting for an entire field to mature?

Use this rough comparison:

adjusted_heat = surprise + compression + mechanism + generativity - adjacency
alpha = leverage + transfer + underpricing + actionability
desire = adjusted_heat * alpha

Score adjacency from 0 to 3 for already-known, already-shipped, or merely renamed ideas. Do not present desire as an objective scientific measurement.

Keep Evidence Orthogonal to Desire

Do not let weak evidence masquerade as alpha, but do not erase a fertile conjecture merely because it is early. Assign an evidence class separately:

• Verified witness: explicit object or result independently checkable.
• Demonstrated: primary artifact with controlled experiments, code, or detailed receipts.
• Suggestive: plausible primary analysis with limited scope or missing replication.
• Claim-only: announcement, anecdote, or unavailable artifact.

Rank verified and demonstrated items in the main list. Put unusually generative claim-only items in a clearly labeled high-voltage sleepers section. Never silently promote them.

Apply the Horny Test

Look for this cognitive sequence:

That cannot be right -> the mechanism is clean -> the evidence bites -> wait, this generalizes.

The final wait is the essential signal: the implication tree has begun branching.

Promote an idea when at least three of these are true:

• It removes an assumption the field treats as structural.
• It compresses several problems into one mechanism.
• It supplies a new control knob rather than another observation.
• It terminates in an object, program, equation, or discriminating experiment.
• It creates an obvious combination with another item in the corpus.
• Its practical or conceptual implication is under-discussed relative to the result.

Kill Anti-Libido

Demote or discard:

• leaderboard gains without a mechanism;
• routine model or product launches;
• enormous claims without inspectable artifacts;
• renamed versions of known objectives;
• scale claims without a causal handle;
• elegant philosophy that produces no discriminating experiment;
• thin demos whose implication depends on ignoring the baseline;
• engagement metrics used as a proxy for intellectual value;
• summaries that are sexier than their source.

Likes, reposts, and bookmark ratios may measure attention or deferred curiosity. Do not confuse them with epistemic alpha.

Write the Ranked Cut

Lead with the outcome and selection contract. Then give a short ranked list, usually 8-15 items.

For each item, provide:

1. Blade: the provocative consequence in one sentence.
2. Mechanism: the smallest explanation that makes the result intelligible.
3. Alpha: what becomes newly possible or newly cheap.
4. Contract: evidence class, scope, and the boundary beyond which the claim is not established.
5. Source: link the primary artifact and, when useful, the captured post.

After the ranking, add only the sections that improve the decision:

• High-voltage sleepers for generative but unverified claims;
• Demotions for popular items that failed the baseline or mechanism test;
• Synthesis for a shared substrate spanning several winners;
• Experiments to steal for the two or three combinations with the highest immediate leverage.

Use charged language when the user asks for it, but anchor every blade with an operational contract. Preserve the heat; expose the receipts.

Final Check

Before finishing, ensure:

• exact artifacts, not titles or virality, determined the ranking;
• novelty was evaluated relative to the user's known baseline;
• mechanism and implication were not conflated with the authors' measured claim;
• evidence status is visible beside every extraordinary assertion;
• the winners generate new work rather than merely reward admiration;
• the final synthesis says something stronger than the individual summaries.