A new Gladstone AI report suggests recursive self‑improvement may kick in within the next 3‑12 months, while Deutsche Bank projects superintelligent AI could outpace human experts by 2030, reshaping competition and markets.
Deutsche Bank Research predicts that by 2030 superintelligent artificial intelligence could outperform human experts across virtually every cognitive task, a shift that would dramatically alter business competition and the global economy. The projection underscores growing investor interest in AI breakthroughs and the strategic importance of preparing for rapid capability gains.
A freshly released report by the analytics firm Gladstone AI, based on interviews with researchers from leading AI labs, paints a more immediate picture. The authors say insiders believe recursive self‑improvement (RSI)—the process by which AI systems autonomously enhance their own design—will begin within the next three to twelve months, provided there is no significant U.S. government intervention or slowdown.
The report outlines how AI companies are automating the entire development pipeline: generating novel research ideas, running experiments to test them, and interpreting results without human oversight. Early‑stage training (pre‑training) focuses on extracting more intelligence per unit of compute, while a portfolio approach allocates limited compute to multiple ideas, scaling up only those that show promise. Because compute resources are finite, the ability to filter good ideas from a flood of AI‑generated proposals becomes a critical bottleneck.
Gladstone’s chief technology officer, Edouard Harris, notes a “fair amount of disaster prepping and bio‑bunker purchasing” among researchers, highlighting the perceived existential risk. CEO Jeremie Harris adds that the goal is to define RSI clearly so policymakers in Washington can craft sensible regulations, rather than blanket bans on superintelligence.
If RSI materializes on the near‑term timeline suggested by the report, the path to the 2030 superintelligence milestone could accelerate dramatically, intensifying calls for robust oversight and international coordination. Stakeholders across industry and government are now urged to monitor these developments closely as the AI landscape edges toward a potentially transformative era.