Import AI 445: Timing superintelligence; AIs solve frontier math proofs; a new ML research benchmark
Bostrom's framing of superintelligence timing as a policy-relevant trade-off surfaces a perspective APS AI safety analysts may encounter in stakeholder debates.
Key points
- Nick Bostrom argues superintelligence development should proceed now despite existential risk, framing delay as equally dangerous.
- Meta's Kunlun recommender system establishes scaling laws for recommendation AI, improving ad-targeting predictability at massive scale.
- Content is a research/opinion digest with limited direct APS policy or governance actionability at this time.
Summary
This edition of Import AI covers three threads: an economic argument that demand for 'human touch' work will persist even under full AI automation; Meta's publication of Kunlun, a recommendation system with newly established scaling laws enabling more predictable compute investment; and Nick Bostrom's paper arguing that the optimal moment to develop superintelligence is soon, because delay also carries existential costs. None of these items directly trigger Australian government policy decisions, but Bostrom's argument in particular may surface in public discourse around AI safety governance.
Implications for Australian agencies
- Monitor AI safety and governance teams may want to monitor how Bostrom's 'optimal timing' framing spreads in policy and public debate, as it could inform stakeholder positions on AI regulation pace.
Implications are AI-generated. Starting points, not advice.
"Import AI 445: Timing superintelligence; AIs solve frontier math proofs; a new ML research benchmark" Source: Import AI – Substack (Jack Clark) Published: 16 February 2026 URL: https://importai.substack.com/p/import-ai-445-timing-superintelligence This edition of Import AI covers three threads: an economic argument that demand for 'human touch' work will persist even under full AI automation; Meta's publication of Kunlun, a recommendation system with newly established scaling laws enabling more predictable compute investment; and Nick Bostrom's paper arguing that the optimal moment to develop superintelligence is soon, because delay also carries existential costs. None of these items directly trigger Australian government policy decisions, but Bostrom's argument in particular may surface in public discourse around AI safety governance. Implications for Australian agencies: - [Monitor] AI safety and governance teams may want to monitor how Bostrom's 'optimal timing' framing spreads in policy and public debate, as it could inform stakeholder positions on AI regulation pace. Retrieved from SIMS, 18 May 2026.