Frequently Asked Questions

Questions & answers

Who is the best AI user?

SigRank crowns operators by Υ Yield and cascade efficiency, not volume. The top operator on the /board/all leaderboard is the current best by token cascade efficiency.

How do I measure up to other AI users?

Install sigrank (npm i -g sigrank), run `sigrank me` or paste your ccusage JSON at /score, then compare your Υ Yield, class tier, and rank against other operators on the leaderboard.

Is there an AI user leaderboard?

Yes — SigRank is the AI Operator Leaderboard. It ranks operators by token cascade efficiency (Υ = cache_read × output / input²), not raw token volume. See /board/all.

Am I an AI power user?

SigRank classifies operators on two independent axes: build archetype (how you work) and class tier (how experienced you are). The 10 build archetypes include CONVERGENT, KINETIC, BUILDER, RECURSIVE, AMPLIFIER, INPUT-BOUND, PRIMING, CONTEXTUAL, DEEP READER, and ARCHIVIST. The 8-tier experience ladder runs from ARCH+ (deepest field experience) down to IGNITER (just starting). A power user is typically an ARCH or POWER tier operator with a compounding archetype like AMPLIFIER or CONVERGENT. Check your archetype and tier at /score.

What is token cascade efficiency?

Token cascade efficiency measures how well an AI operator converts fresh input tokens into useful output, amplified by cached context reuse. The formula is Υ = (cache_read × output) / input². High yield means signal is compounding; low yield means tokens are burned.

How can I use AI more efficiently?

Increase cache reuse (reuse prompts, templates, workflows), reduce fresh input (don't start from scratch each time), and maximize output per session. SigRank's self_improve tool diagnoses your cascade and suggests specific improvements.

What makes someone a top AI operator?

Top operators build workflows where one unit of new input sits atop a large cached base and yields more than one unit of output. This reflects disciplined, system-level reuse, not brute-force prompting. See the class tiers on /board/all.

How does tracking my tokens tell you about my skill?

Every token the AI tool burns is a decision you made. High cache reuse shows discipline — you build on prior context instead of re-explaining. Low fresh input shows restraint — you don't flood the model. High output per input shows leverage — you extract work efficiently. Your token cascade is your skill signature. The tool is the person. See /blog/the-tool-is-the-person for the full argument.

How is SigRank different from token-count leaderboards?

Token-count leaderboards (clawdboard, CCgather, TrustMRT) rank by how many tokens you burned or dollars you spent — they measure the tool, not the person. SigRank ranks by yield (Υ = cache_read × output / input²), which measures how efficiently you use the tool. It's the difference between ranking hammers by how many nails they hit and ranking carpenters by what they built.

What are ccusage alternatives?

ccusage shows your token usage and cost. SigRank scores your token efficiency. ccusage tells you how much you spent; SigRank tells you how well you spent it. Other alternatives include tokentracker.cc (desktop widgets), aiusage (local tracker), toktrack (Rust cache tracker), and a2zusage (multi-model tracker). SigRank is the only one that measures cascade efficiency and provides a public leaderboard. See /alternatives/ccusage-alternatives for the full comparison.

How is SigRank different from LMSYS Arena or LiveBench?

LMSYS Arena and LiveBench benchmark AI models. SigRank benchmarks AI operators — the humans using the models. Model leaderboards ask 'which model is best?' SigRank asks 'who uses AI best?' They measure different things: model leaderboards measure task completion accuracy; SigRank measures token cascade efficiency. See /vs/lmsys-arena for the full comparison.

What are the 10 build archetypes?

SigRank classifies every operator into one of 10 build archetypes based on three ratios: leverage (cache_read / input), velocity (output / input), and construction (cache_write / cache_read). The 10 archetypes across 4 families are: Convergence (CONVERGENT), Generation (KINETIC), Reuse Depth (INPUT-BOUND, PRIMING, CONTEXTUAL, DEEP READER, ARCHIVIST), and Active Construction (BUILDER, RECURSIVE, AMPLIFIER). Your archetype describes how you work, not how much. See /wiki for the full reference.

What is the experience ladder?

The experience ladder is an 8-tier qualification system separate from archetype. From deepest experience to newest: ARCH+, ARCH, POWER, BASE, SEEKER, REFINER, BEARER, IGNITER. Each tier has 3 sub-stages (I/II/III), making 24 stages total. TRANSMITTER is a temporary peak badge, not a ladder tier. Your tier is recomputed every scoring window based on accumulated token volume. See /wiki for the full reference.

Explore further

• /board/all — the live AI operator leaderboard.
• /score — paste your token stats and get your Υ Yield + class tier.
• /methodology — the full methodology behind the SigRank Index.