AI makes answers abundant.
TQ measures what you do next.
Get a problem you have never seen. Think, work with AI, verify, decide, and act. TQ measures how you solve — then shows you how to get better.
One session. One trajectory. One useful result.
See the problem
Start before AI. Define what is known, what is missing, and what actually needs to be solved.
Work with AI
Ask, challenge, verify, reframe, and use the model as a collaborator rather than an answer machine.
Leave with a move
Finish with a concrete solution and a first real-world action. TQ scores the path that got you there.
Discovery · Framing · Decomposition · AI Collaboration · Verification · Reasoning · Adaptation · Execution · Learning Velocity
Not an IQ test. Not a prompt-writing contest. Not a benchmark you can fake with a polished final answer.
TQ Playground
No topic menu. No level selector. TQ gives you a new real-world problem every session so the task stays unfamiliar.
Ready for a new problem?
TQ chooses the problem, constraints, and difficulty for you. You just start.
What would you do?
Explain your approach, key assumptions, what you would check, and the first action you would take.
How TQ is recording this session
Your result
Practice score — useful for tracking your growth.
See capability breakdown & trajectory
Capability dimensions
Your trajectory
What changed your score
Your response
What a strong solution could look like
How to improve
See evaluation rubric
Use this to decide what to improve next.
See where your TQ comes from
Your score is only useful with its shape: nine capabilities across Discover, Solve, and Execute.
TQ Capability Model v1 — nine dimensions
Compare your early sessions with your latest ones to see growth.
Your TQ (provisional)
Growth since your first sessions
Run at least two sessions to see growth.
NEXT CHEAPEST POINT GAIN
Your graph needs data. Finish a challenge and the weakest dimension becomes obvious.
Training dashboard
Your recent practice, streak, AI use, and progress.
This week
Your contribution to the Human × AI dataset
Your data stays in this browser in the current demo.
Your next best training move
Coach uses your past sessions to pick the skill worth training next.
Where your capability is leaking
This week’s plan
The plan updates from your recent results.
Did your score predict a real result?
Log what you actually did after a session, then compare the result with your TQ. This is how the score earns meaning.
Log a real-world outcome
Does the score predict the outcome?
With n < 30 any correlation is noise. That is exactly why the loop exists — log outcomes after real work, not hypotheticals.
Recent outcomes
Nothing logged yet.
100 people enjoy unfamiliar problems
Proof that people want to train this way — the gym, not the exam, is the product.
1,000 scored sessions
Proof that Human × AI behavior can be measured consistently — 100 learners × 10 problems.
100+ real-world outcomes
The one that matters: proof that TQ predicts real AI-assisted performance. This is where serious investors start looking.
Measure capability after practice
Three timed challenges using the same scoring engine. Practice first; the assessment measures trained capability.
Latest assessment report
No assessment on record. Reports include TQ, band, nine dimensions with confidence, and the trajectory evidence behind each.
Why assessment comes after the gym
A one-shot exam rewards people who test well. The gym builds the capability; the assessment measures a stable profile; validation checks the profile predicts reality. Each layer only means something because of the one before it:
A credential is a promise about prediction
A credential only matters when the score has survived formal assessment and real-world validation.
TQ Capability Credential
Issued only after practice, formal assessment and validation. Use “Download / Print PDF” to save it as a landscape PDF.
Free to train. Pro to grow. Workforce to decide.
The gym is free forever — it is where users and data come from. Pro is the depth on top of your own data. Workforce is what enterprises actually want to buy: not “he says he uses AI”, but a validated capability profile. We do not start with enterprise sales; the first hundred-thousand users are the ones AI left confused about what they themselves can now do.
TQ Gym
- Unlimited challenges · all 7 domains · L1–L3
- Full trajectory recording & session reports
- Capability graph (current + best)
- Daily challenge & weekly plan
- Validation-loop outcome logging
TQ Pro
- Full capability graph history & growth analytics
- Advanced constrained challenges (L3+, custom briefs)
- AI coach: deep trajectory review, move by move
- Progress reports & personal benchmarking
- Priority assessment windows · full reports
TQ for Organizations
- Team capability profiles (consent-based, aggregate-first)
- Hiring / promotion / reskilling signals
- Program analytics & benchmark by department
- Credential verification endpoint
Who we build for first — people AI hasn’t given a map yet
Their problem is not “I can’t use AI”. It is “I don’t know what AI now makes possible for me.” The gym answers that in ten minutes; enterprise sales come after the score predicts something.
Workforce preview · aggregate by default
member:read_pii scope gate, and consent-based individual data.Capability vs market benchmark
Credential funnel
Department scorecard
| Dept | Head | Part. | Avg TQ | Δ |
|---|
The data layer, in public
The dataset is the asset; transparency is how it stays trustworthy. Everything TQ has recorded from you, the scoring model’s calibration status, and your export button — one place.
Scoring model · TQ Capability Model v1
| Dimension | n | α | Discrim. D | Grade |
|---|
α (internal consistency) > .70 acceptable · D = pupper − plower (top/bottom 27%) ≥ .40 excellent. Computed from your sessions only — real calibration needs the network.
Your dataset · export
Feature flags
One identity across every layer
Your TQID links your gym sessions, capability graph, assessment reports, validation outcomes and — eventually — your credential.
Set your display name
A TQ Identity is created automatically on your first session — this step is optional and sets the name on your assessment report and credential. TQ is measured against your own progress — never ranked against other people.
localStorage): identity, sessions, trajectories, outcomes. No server, no upload. Clearing site data resets the demo.