🔍 Read the full analysis: From AI To AGI And ASI: What 50 Superintelligence Signs Suggest on ThorstenMeyerAI.com
Get business pricing on office and shipping supplies
- Business-only prices and quantity discounts
- Tax-exempt purchasing
- Multiple users, one account, clear invoices
TL;DR
Australian AI consultant Alan D. Thompson’s 50-indicator tracker lists two signs as achieved, 33 as partially achieved and 15 as not achieved. A review of 18 prominent claims found seven verified, five based on company claims, three resting on one source and three disputed or contradicted; the tracker is one researcher’s framework, not a standard measure of superintelligence.
The tracker, published by Australian AI consultant **Alan D. Thompson**, ranges from advances in mathematics to ambitions such as colonizing another planet. In an October 26 review, ThorstenMeyerAI.com traced 18 selected claims to papers, lab posts, researchers or official documents. It classified **seven as verified, five as vendor claims, three as based on a single source, and three as disputed or contradicted**. Those categories describe the selected claims, not all 50 indicators.
The tracker’s two achieved entries concern a **major mathematical proof and a resolved conjecture**. Several mathematical claims reviewed had support from named researchers or external mathematicians, although the accounts include qualifications. For example, mathematician Terence Tao described work on an Erdős problem as “more or less autonomously” completed after feedback; the problem had been misformulated, and similar results already existed. Donald Knuth described Claude finding a construction after coached exploration, while Knuth wrote the proof.
Thompson’s meaning of “partially achieved” is broad. An August correction note says orange status can include **funded programmes, research infrastructure, formal scientific missions and enabling work**. The review argues that this does not mean an indicator is close to completion. It also reports that two exoplanet candidates identified with Claude Code were associated with an indicator about colonizing a planet, illustrating how a tracker entry can cover steps far short of the stated outcome.
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
Why the ASI label depends on the scale
The tracker is a **watchlist assembled by one researcher**, not an agreed scientific test for superintelligence. Readers may otherwise mistake a partially achieved status for evidence that a system is near a milestone, or treat the label “early ASI” as a finding shared by AI labs and researchers.
Definitions materially change the picture. Thompson defines ASI as expert-human performance in practically any field. Google DeepMind’s 2023 framework calls performance at the expert level **“Expert AGI”** and distinguishes higher levels, including superhuman performance. Its June 2026 report describes superintelligence as exceeding the cognitive capability of large human organisations. OpenAI’s charter, meanwhile, defines AGI around highly autonomous systems outperforming humans at most economically valuable work. Those standards focus on different capabilities, so a strong result in mathematics does not by itself establish broad, autonomous superiority.
The review also points to a gap between impressive demonstrations and systems that can replace or outperform people across real work. It reports no household robots operating at scale, no verifiably billion-dollar AI-run company, and no visible break in the US unemployment rate. These are specific observations from the source, not proof that AI has had no economic effect.
Products Worth Considering
HP Sprocket Photo Booth Machine Instant Color Photo Printer with 10.1 Touchscreen, LED Ring Light, 250+ Frames, 3×4 ZINK Paper, QR Code Shar
Dell UltraSharp 32 U3226QER 4K Thunderbolt Hub Monitor
Apple iPad Air (2025) 13‑inch Wi‑Fi 256 GB – Space Gray (M3 chip)
As an affiliate, we earn on qualifying purchases.
From a checklist to competing definitions
Thompson’s list contains **50 indicators**, including recursive code optimization and much more speculative milestones. His October video, “ASI: A better world – 2026,” presents entries from January through October. The source describes Claude Mythos Preview and GPT-6 Astra as systems Thompson calls “early ASI”; that characterization is his assessment, not a consensus classification.
The source says Thompson also maintains a separate countdown to AGI at 99%. It reports that he reserves the final percentage for a robot able to perform multi-step fine assembly, such as assembling flat-pack furniture, with average-human reliability. That countdown and its chosen threshold are likewise **Thompson’s framework**, rather than a field-wide benchmark.
Evidence quality varies by claim. The review treated independent expert assessment or official documentation as verification, while company figures without independent reproduction were labeled vendor claims. It notes that public formal proof certificates can support mathematical results, but that human contributions, coaching and the choice of problem still matter when interpreting what a model accomplished.
“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
What the tracker cannot establish
The **50 indicators are not a standardized measurement**, and the source review checks only 18 selected claims. It does not establish that the remaining entries are true or false, nor does it provide a common method for comparing progress across them. Some claims rely on a single source, while others are company-reported or disputed.
It is also unclear from the source how the tracker’s status judgments are independently audited or how much weight each indicator carries. A result in a narrow domain may show real capability without demonstrating broad general performance, autonomy or the ability to exceed human organisations. The source’s examples do not settle whether any current model meets a shared definition of AGI or ASI.
Evidence to watch in future updates
The next useful developments will be **independent replication** of company-reported results, clear documentation of human guidance and tool use, and tests that measure performance across domains rather than a single benchmark. Public proof certificates and peer review can help assess mathematical claims, but broader claims about autonomy and real-world work require evidence suited to those tasks.
Thompson’s tracker may change as new claims are added or statuses are revised. Readers should compare any update with the underlying source and its definition of success, rather than relying on the color or label alone. The October review does not report a forthcoming independent audit or a settled date for a field-wide ASI standard.
Key Questions
How many of the 50 signs does the tracker mark achieved?
As reported for **October 26, 2026**, Thompson’s tracker marks two indicators achieved, 33 partially achieved and 15 not achieved.
Does the tracker show that artificial superintelligence has arrived?
No. It records one researcher’s assessments, and its **partially achieved** category includes funded work and other enabling activity. The review says the tracker should be read as a watchlist, not a measurement accepted across the field.
Why do researchers disagree about what counts as ASI?
They use different thresholds. Thompson’s definition emphasizes expert-level performance across fields, while Google DeepMind’s 2026 description compares superintelligence with the capabilities of **large organisations of humans**. Other frameworks emphasize autonomy and economically valuable work.
Were the mathematical AI claims independently checked?
Some received support from external mathematicians or published work, according to the review. Others involved **coaching, feedback or human-written proofs**, so a verified result does not necessarily mean a model solved a problem entirely on its own.
What evidence would strengthen claims of ASI?
Independent replication, transparent accounts of human involvement, and reliable evaluations across a wide range of tasks would help. A narrow technical result alone does not establish broad **autonomy or superiority over human organisations**.
Source: ThorstenMeyerAI.com
