The Case for the Name Super Intelligence
We support Executive Order 14434. President Trump signed the order on September 29, 2026, directing federal agencies to replace “Artificial Intelligence” with “Super Intelligence” (Source: Federal Register), and Iternal backs that name on the merits: the results are real, the performance is measured, and “Super Intelligence” finally describes what this technology does instead of how it was built.
The Name Now Describes the Capability, Not the Origin
The name is older than every result it now describes. John McCarthy and his co-authors coined “artificial intelligence” in a 1955 proposal for a two-month Dartmouth summer study, betting that a machine “can be made to simulate” intelligence (Source: AI Magazine, AAAI reprint). “Simulate” fit 1955, because nothing had shipped yet. It does not fit 2026, when the results are measured and real.
The order says the same in its own text: these capabilities “increasingly represent not merely artificial intelligence, but a new era of Super Intelligence” (Source: Executive Order 14434, Sec. 1).
+15% more issues resolved per hour
Across 5,172 customer-support agents, generative AI raised issues resolved per hour by 15% on average — about 30% for less-skilled and less-experienced agents (Source: Brynjolfsson, Li & Raymond, Quarterly Journal of Economics).
+26% more weekly tasks completed
Across 4,867 developers at Microsoft, Accenture and a Fortune 100 electronics manufacturer, AI coding assistance lifted completed weekly tasks by 26.08% (Source: Cui, Demirer, Jaffe, Musolff, Peng & Salz, Management Science).
80% report a productivity lift
Across 1,719 enterprise respondents in 97 nations, 80% said AI improved their individual productivity (Source: McKinsey, “The State of AI in 2026”).
Two of these are counted outputs from real workplaces, published in peer-reviewed journals; the third is what 1,719 enterprise respondents report about their own work. None of them is a simulation. Results like these are why “artificial” stopped being the accurate word for what this technology does.
The Performance Clears the Human Bar, Certified and Leaderboard-Verified
The strongest evidence for the name does not depend on an IQ test at all. It is certified competition results and leaderboard-verified benchmark scores, measured against a known human baseline:
| Benchmark | Leading model | Human comparison |
|---|---|---|
| IMO 2025, officially certified | 35 of 42 points — gold-medal standard, solving 5 of 6 problems in natural language (Source: Google DeepMind) | 72 of 630 contestants earned gold that year (cutoff ≥35) (Source: imo-official.org) |
| ICPC 2025 World Finals | 12 of 12 problems solved (OpenAI’s system) (Source: Mostafa Rohaninejad via Simon Willison); Google DeepMind’s Gemini 2.5 Deep Think solved 10 of 12, gold-medal level, including one problem no university team solved (Source: Google DeepMind) | Best human team, St. Petersburg State University: 11 of 12 (Source: cphof.org standings) |
| GPQA Diamond (graduate-level science) | 96.3% (Artificial Analysis) (Source: Artificial Analysis) / 95.8% ±1.4% (Epoch AI) (Source: Epoch AI) | 81.3% — the original PhD-expert validator baseline, a figure the paper itself says is skewed upward by how the questions were selected (Source: Rein et al., arXiv:2311.12022) |
| ARC-AGI-2 (abstract reasoning) | 95.0% on the Foundation’s own leaderboard (Source: ARC Prize leaderboard) | 60% average human test-taker (Source: ARC Prize blog) (66% of attempted test pairs, per the Foundation’s later paper) (Source: arXiv:2505.11831); a panel of humans collectively solves 100% |
IQ scores are a second, dated layer behind the table above, not the first: the certified reasoning and science results already carry the case on their own. The IQ figures, how TrackingAI.org tracks them weekly, and the full position — every pillar, every citation — are in Why Iternal Supports Super Intelligence.