🔍 Read the full analysis: What Do 50 Superintelligence Signs Reveal About AI’s Future? on ThorstenMeyerAI.com
Get hardware and tech essentials delivered free with Prime
- Fast, free delivery on millions of items
- Prime Video, Amazon Music and more included
- Member-only deals all year
TL;DR
AI consultant Alan D. Thompson’s tracker marks two of 50 proposed signs of artificial superintelligence achieved, 33 partially achieved and 15 not achieved. A review of 18 prominent claims found verified results in mathematics and science, alongside vendor-reported results and disputed claims; the tracker is an individual framework, not a field-wide measurement.
AI consultant Alan D. Thompson’s tracker lists two of 50 indicators of artificial superintelligence as achieved, 33 as partially achieved and 15 as not achieved, according to its status grid updated October 26. A review of 18 prominent claims in his 2026 video found verified AI contributions in areas including mathematics, but also vendor claims, single-source reports and disputed findings — evidence of advances, not an independent determination that superintelligence has arrived.
Thompson, an Australian AI consultant who publishes The Memo newsletter, maintains a public checklist ranging from recursive code optimization to much more speculative milestones. In his tracker, he describes models including Claude Mythos Preview and GPT-6 Astra as “early ASI.” That label reflects his own definition and judgment; it is not a consensus classification shared by AI labs or researchers.
The October 26 tally is two achieved, 33 partially achieved and 15 not achieved. Thompson’s August correction note says “partially” includes documented progress such as funded programs, research infrastructure, scientific missions and enabling work. It does not mean an indicator is almost complete. For example, the source says two exoplanet candidates found using Claude Code are filed under a tracker item about colonizing a planet other than Earth, a categorization that illustrates how broadly some indicators can be interpreted.
In its review of 18 headline claims, ThorstenMeyerAI.com reports seven verified, five vendor claims, three based on a single source, and three disputed or contradicted. The review describes the verified claims as concentrated in areas where results can be examined, including mathematical proofs, a physics calculation and regulatory filings. Its assessment is a source review, not a comprehensive independent audit of every tracker entry.
Whose superintelligence?
One word, very different bars
Dr Alan D. Thompson's ASI starts where Google DeepMind's scale puts "Expert AGI". The same model can be "early ASI" on one scale and a strong but narrow AGI candidate on another.
- Median human50th percentileThompson: AGIa machine at the level of an average (median) humanDeepMind: Competent AGIat least the 50th percentile of skilled adults
- Expert90th percentileThompson: ASIexpert-human level in practically any fieldDeepMind: Expert AGIat least the 90th percentile of skilled adults
- Virtuoso99th percentile—DeepMind: Virtuoso AGIat least the 99th percentile of skilled adults
- Beyond every human100%—DeepMind: Superhuman (Level 5)outperforms 100% of humans
- Beyond organisations2026 report—DeepMind: ASImore capable than large organisations of humans
Thompson’s tracker status as of 26 October 2026. DeepMind rows: “Levels of AGI” (2023) and, for the last row, “From AGI to ASI” (June 2026). OpenAI’s Charter: AGI = “highly autonomous systems that outperform humans at most economically valuable work”.
Sources: LifeArchitect.ai ASI checklist (26 Oct 2026), AGI countdown, Morris et al., Levels of AGI (2023), Genewein et al., From AGI to ASI (2026), OpenAI Charter.
Powered by Thorsten Meyer AI
Evidence Is Strongest in Verifiable Tasks
The findings matter because declarations that AI has reached “superintelligence” depend heavily on what counts as intelligence and how independently a result can be checked. A model producing a proof or helping solve a research problem may show genuine capability. That does not by itself establish broad, autonomous performance across fields, or demonstrate that a system can reliably perform work at the scale implied by superintelligence.
The review’s strongest examples are notable precisely because experts or public records can scrutinize them. Mathematicians Terence Tao, Donald Knuth and Timothy Gowers are cited as describing substantive AI contributions, while acknowledging qualifications involving human guidance, existing literature or the style of the problem. These cases support a measured conclusion: AI can contribute to advanced research, but the degree of independence and generality varies by task.
Other proposed signs concern economic and physical-world outcomes, which are harder to infer from demonstrations or research programs. The source reports no household robots operating at scale, no verifiably billion-dollar company run by AI, and no visible break in the US unemployment rate attributed to AI. Those observations do not prove such changes cannot happen; they caution against treating research progress as evidence that society-wide effects have already arrived.
Products Worth Considering
Ledger Stax Crypto Wallet with Touchscreen
Trezor Safe 5 Crypto Hardware Wallet with Color Touchscreen, Black Graphite
Apple MacBook Pro 2023 14 with M3 Pro Chip, 18 GB RAM, 2 TB SSD, Space Black
As an affiliate, we earn on qualifying purchases.
Definitions Change the Threshold
Thompson defines AGI as a machine performing at the level of an average human and ASI as one performing at the level of an expert human in practically any field. Google DeepMind’s 2023 “Levels of AGI” framework uses different categories: competent, expert, virtuoso and superhuman. Under that scale, expert performance is below the superhuman level, so Thompson’s ASI label does not map directly to the terminology used by DeepMind.
DeepMind’s June 2026 report describes superintelligence as a system more intelligent and cognitively capable than large organizations of people. OpenAI’s Charter, meanwhile, defines AGI around highly autonomous systems that outperform humans at most economically valuable work. These definitions emphasize different things — expert skill, performance beyond humans, or autonomy — and can produce different judgments about the same model.
Thompson also maintains a separate “conservative countdown to AGI,” which the source says stood at 99%. He identifies reliable, multi-step fine assembly by a robot as the remaining hurdle on that measure. That figure is a feature of his own tracker, not a general scientific estimate of how close AGI is.
“documented material progress toward an indicator, including funded programs, research infrastructure, formal scientific missions, and enabling work”
— Alan D. Thompson, in an August correction note about the tracker
The Meaning of Partial Progress
The tracker does not establish that its 50 indicators are a validated measure of superintelligence. Its author’s definitions and status judgments shape the results, and the source review covers 18 selected claims, not every item. The available material also does not provide enough detail to independently assess all 18 verdicts or reproduce every claimed result.
For several highlighted developments, the distinction between a model’s contribution and the work of people remains material. The source describes mathematical achievements involving coaching, human-written proofs or problem formulations that needed correction. It also labels some results as vendor claims, meaning the figures come from companies and have not been independently reproduced according to the review. The source’s excerpt does not include full details for every example or the methods behind each assessment.
It is also unclear how much of the reported progress generalizes beyond specific benchmarks and research tasks, or whether it produces reliable autonomy in real-world settings. The tracker’s colored categories should not be read as a forecast of when a milestone will be met. No independent measurement in the supplied material settles whether current systems meet a commonly accepted definition of AGI or ASI.
More Replication, Broader Tests
The next useful evidence will be independent replication of claimed results, clearer reporting of the human input involved, and tests that distinguish performance on a narrow problem from reliable work across different fields. Public proof certificates or official records can help reviewers check particular outcomes, but they do not by themselves answer questions about autonomy or broad capability.
Readers following Thompson’s tracker can watch for changes to its status grid and compare its entries with assessments from AI labs and independent researchers. The source does not identify a scheduled independent review of the full list or a forthcoming event that would settle the debate. For now, the clearest conclusion is that some AI research contributions are documented, while the broader claim that superintelligence has arrived remains dependent on definitions and evidence that have not been established across the field.
Key Questions
How many of the 50 indicators are marked achieved?
Thompson’s tracker, updated October 26, 2026, marks two achieved, 33 partially achieved and 15 not achieved.
Does “partially achieved” mean a sign is nearly complete?
No. Thompson says the category can include funded programs, research infrastructure, formal missions and other enabling work. It does not necessarily mean the indicator is close to being met.
Has the review established that AI is superintelligent?
No. The source review found evidence for particular AI contributions, but it did not establish that AI systems meet a shared definition of superintelligence. The conclusion depends in part on definitions and on how independently results are verified.
Why do researchers use different labels for AI capability?
Frameworks set different thresholds. Thompson defines ASI around expert-human performance in practically any field, while Google DeepMind’s description places superintelligence beyond the capabilities of large human organizations. Labels such as AGI and ASI therefore do not always refer to the same level.
Source: ThorstenMeyerAI.com
