Not: Forecasts
Our simulations are scenarios under stated assumptions. We report odds across many runs, never a prediction of the future.
About
CUSP42 Labs studies two questions at the same depth: what could go wrong as AI grows more capable, and how deeply AI will reshape work, institutions and what it means to be human. We measure, simulate, debate and imagine, and we publish our assumptions.
Why
Public debate about AI swings between fear and promotion, and both usually skip the mechanisms. A thought lab can do something slower and more useful: make the assumptions explicit, run them, show the spread of outcomes, and let people change the inputs. The transition is too important to argue about with slogans or one model's fluent certainty.
What we do not claim
An AI-risk lab is only as credible as its honesty about what its work can and cannot show.
Our simulations are scenarios under stated assumptions. We report odds across many runs, never a prediction of the future.
We do not rank countries by their political systems or assign blame for wars. One set of rules applies to every country.
CARI is a structured judgement with written anchors, not a measurement of nature. Its anchors are published so it can be argued with.
We are not an AI-alignment lab. We study AI risk and societal impact through measurement, simulation and analysis.
The name
A cusp is the threshold where a curve changes direction and old assumptions stop working: the AI transition is one. 42 is the famous "ultimate answer" from The Hitchhiker's Guide to the Galaxy, a reminder that the answer is useless without the right question. The lab exists to ask better questions at the threshold.
Independence
CUSP42 Labs is independent. It is not affiliated with, endorsed by, or representative of any employer, institution, client, development bank, government or official organisation. Its tools are free and its methods are published. To cite: CUSP42 Labs (2026), with the tool's name and address; see each tool's page.