Bets that resolve.
Public predictions, dated, with explicit confidence and resolution conditions. Some are about the industry. Some are about me. All of them will be marked won, lost, or void in time. A wager is the cleanest unit of accountability for a public thinker. This is the page that calls bullshit on the rest of the site.
Live bets
-
A meaningful US bill (meaning one that gets a committee vote) will frame personal data as a property right, not a privacy right, before the end of 2026.
Privacy frameworks have failed to produce payable cash flow. The legislative bottleneck is who collects, not who hides.
posted 2026-04-15 #Human data rights -
I will release a six-track Abhi Cris record before end of year, built around tempo as a single argument across all six.
The skeleton is written; mixes for three tracks are 70% in. Public commitment is the last forcing function the work needs.
posted 2026-05-06 #Music as system -
Two of the five highest-funded ‘creator economy’ networks of the 2021–2023 vintage will sunset or quietly pivot off creators by mid-2027.
Distribution-only networks were always renting their growth from another network. Generation getting cheap removes their last input cost moat.
posted 2026-03-10 #Decentralized creative economy -
By end of 2027, at least one major AI lab will route inference through a programmatic licensing protocol: not as PR, but as architecture.
Litigation cost is now exceeding what a per-call licensing fee would have been. The first lab to admit it publicly wins the policy narrative.
posted 2026-05-01 #IP economics
How these resolve
- Every wager has a posted date, a resolution date, and a confidence level, set in advance, never revised.
- A bet is won if the claim is plainly true at the resolution date. Lost if plainly false. Void only if the world materially changed in a way the claim could not contemplate (e.g. the entity referenced no longer exists).
- Resolutions are written publicly, with one paragraph of post-mortem. Wrong is fine; vague is not.
- The point of a wager is not to be right. It is to be the kind of person whose claims are checkable.