Philanthropy's urgent opportunity to create the interim international AI institution
Bibliographic record
Abstract
The rapid development of new artificial intelligence (AI) technologies has outpaced the ability of regulators in most parts of the world to put rules in place that govern their use. Without regulation of AI, its benefits are likely to flow to the few, while the many risks it poses and the harms it has already wrought will be borne by society, and disproportionately so by already vulnerable communities. Many of AI's problems are inherently global, which means that if regulation takes place in a loose patchwork, solutions will be evasive: certain AI products, services, practices or tools outlawed in one part of the world may still be available through companies located in other jurisdictions. David Evan Harris and Anamitra Deb argue that in the context of the urgent need for truly meaningful regulation of AI, philanthropy has an opportunity to quickly leverage its stores of uniquely public-interest-bound "risk capital" to create the Interim International AI Institution. While numerous efforts are under way to start conversations and study what a possible international or intergovernmental AI governance body might look like, this proposal suggests simply putting forth the funding and prototyping the organization by beginning the work today. The critically important work of this institution would include coordinating conversations among governments around the world that are now developing potentially incompatible AI governance regulations in parallel; establishing best practices and norms for AI governance; bringing together a critical mass of technical, legal, policy and social science expertise; and transparently sharing the fruits of its rapid and iterative AI governance prototyping efforts.
Fetched live from OpenAlex and de-inverted. Abstracts are not stored in this database: the inverted indexes are 8.6 GB of the frame’s 9.3 GB of text, and the host has 13 GB free.
How this classification was reachedexpand
Full frame machine prediction
Teacher imitationNot calibrated prevalence, not ground truth. Human validation pending. The Gemma side is a direct model label for every work in the frame, read from the title-only record. The Codex side is a classifier learned from the 10,348 direct Codex labels and calibrated to design-weighted sample rates; fields without enough sample support carry no Codex call. Candidate is the union of the two sides; consensus is their intersection. These outputs are machine_predicted_unvalidated and are not human labels.
Distilled classifier scores by category (both heads)
| Category | Codex | Gemma |
|---|---|---|
| Metaresearch | 0.028 | 0.026 |
| Meta-epidemiology (narrow) | 0.001 | 0.001 |
| Meta-epidemiology (broad) | 0.001 | 0.001 |
| Bibliometrics | 0.001 | 0.001 |
| Science and technology studies | 0.008 | 0.010 |
| Scholarly communication | 0.024 | 0.030 |
| Open science | 0.002 | 0.023 |
| Research integrity | 0.011 | 0.021 |
| Insufficient payload (model declined to judge) | 0.030 | 0.009 |
Machine scores (provisional)
The two teacher heads of the student model, read on this work. A score orders the frame for review; it never asserts a category, and the validation status ships verbatim with every row.
Baseline scores from an immature model (maturity gate not passed, 7 training rounds). Scores rank; they never assert a category.
score_only:v0-immature-baseline · verbatim from the scoring run: score_only means the number may rank works, and no category label ships from itClassification
machine, unvalidatedMachine predicted; a candidate call from one source (direct Gemma or distilled Codex), not a consensus.
How this classification was reached, model by model and score by score, is at the end of the page under "How this classification was reached".