🔍 Read the full analysis: A Question Of Distance: 50 Signs Of AI Superintelligence, AGI, And ASI on ThorstenMeyerAI.com
Get business pricing on monitors, keyboards and dev gear
- Business-only prices and quantity discounts
- Tax-exempt purchasing
- Multiple users, one account, clear invoices
TL;DR
A report published by ThorstenMeyerAI.com on Oct. 26, 2026, checks 18 prominent claims from researcher Alan D. Thompson’s 50-sign artificial superintelligence tracker. The tracker lists two signs as achieved, 33 as partially achieved and 15 as not achieved, but its definitions are broad and several headline claims rely on company statements or limited evidence.
ThorstenMeyerAI.com published a review on Oct. 26 of 18 prominent claims from Australian AI consultant Alan D. Thompson’s 50-sign artificial superintelligence tracker, finding that some are backed by independent experts or official records while others remain company claims, single-source reports or disputed. The review matters because Thompson describes current leading models as “early ASI,” but the tracker’s labels and definitions are not a standardized measure of machine intelligence.
Thompson’s tracker listed two of 50 indicators as achieved, 33 as partially achieved and 15 as not achieved on Oct. 26. The items range from mathematical results and code optimization to broad milestones such as household robots and human settlement beyond Earth. Meyer’s review says the “partially achieved” category is expansive: it can include funded programmes, research infrastructure, formal missions and other enabling work, rather than a capability that is nearly complete.
Of the 18 headline claims Meyer checked, seven were classified as verified, five as vendor claims, three as relying on a single source, and three as disputed or contradicted. The report says the stronger verification tends to involve results that can be independently examined, including mathematical proofs, a physics calculation and regulatory filings. It does not treat a lab’s own performance figures as independent confirmation.
Mathematics is among the clearest areas of reported progress, though the accounts carry caveats. Meyer cites mathematicians Terence Tao, Donald Knuth and Timothy Gowers as describing genuine AI contributions. The examples still involved human feedback, coaching, literature review or limits tied to the problem, according to the review. It also notes that some publicized proof outputs were checked or completed by people, rather than produced autonomously from start to finish.
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 is contested
The dispute is not only about whether models can solve difficult tasks. It is also about what counts as AGI or ASI, and how much autonomy a system must show. Thompson defines artificial general intelligence as performance at the level of an average person and ASI as expert-human performance in practically any field. Under that definition, strong results in specialized work can be presented as evidence of early ASI.
Other organizations set different bars. Google DeepMind’s 2023 “Levels of AGI” framework distinguishes competent, expert, virtuoso and superhuman performance. Its June 2026 report describes superintelligence as exceeding the cognitive capability of large human organizations. OpenAI’s charter, meanwhile, defines AGI in terms of highly autonomous systems outperforming humans at most economically valuable work. These definitions make claims about “early ASI” difficult to compare unless the speaker’s scale is stated.
For readers, the distinction affects how to interpret claims about jobs, scientific discovery and AI systems improving their own development. A striking result on one carefully chosen task does not by itself establish broad competence, reliable autonomy or an economy-wide impact. Meyer’s review points to important missing evidence: it says there are not yet household robots operating at scale, no verifiably billion-dollar AI-run company, and no visible break in the U.S. unemployment rate attributable to AI.
Products Worth Considering
Jackery Solar Generator 1000 V2 with 200W Solar Panel
Amazon Kindle Paperwhite Signature Edition (12th Generation, 2024), 32 GB, Metallic Black
Bose SoundLink Flex Bluetooth Speaker (2nd Gen), Portable Waterproof Speaker, Black
As an affiliate, we earn on qualifying purchases.
How the tracker defines progress
Thompson’s public checklist contains 50 indicators, including recursive code self-optimization and milestones extending well beyond current AI deployments. His October video, “ASI: A better world – 2026,” presents examples from January through October. The tracker identifies Claude Mythos Preview and GPT-6 Astra as “early ASI” systems; that characterization is Thompson’s assessment, not a consensus finding established by the review.
In an August correction note cited by Meyer, Thompson explained that orange, or partial, status covers “documented material progress” that can include funded programmes and enabling work. One example in the review is two exoplanet candidates reportedly found with Claude Code, which the tracker places under an indicator about colonizing a planet beyond Earth. That classification illustrates why partial status should not be read as evidence that the stated milestone has been accomplished.
Thompson also maintains a separate countdown to AGI that stood at 99%. He says the remaining step depends on a robot completing multi-step fine assembly, such as building flat-pack furniture, with average-human reliability. That countdown is a separate, author-defined estimate, not an independently validated probability that AGI is imminent.
““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 on his tracker, as quoted by Thorsten Meyer
What the evidence cannot settle
The review is a check of 18 selected claims, not a full independent audit of all 50 indicators or every model mentioned. Its classifications depend on available papers, expert comments, company posts and official records. A verified result in a narrow domain does not establish general intelligence, while a vendor claim may be accurate without having been independently reproduced.
The supplied material does not provide the full evidence or detailed verdict for every item in the 18-claim sample. It also does not establish that the named models meet a shared AGI or ASI benchmark. The tracker’s status categories, the model makers’ reported results and the review’s own verification labels should therefore be treated as distinct things. Whether AI has accelerated research overall, or can reliably perform complex work without human direction, remains unsettled by the evidence described.
Evidence to watch in future updates
The next useful signals will be results that outside researchers can reproduce, clearly documented human involvement, and evaluations that measure performance across fields rather than on a single selected task. For claims about autonomy, evidence would need to show sustained work with limited supervision, not just a strong result after coaching or feedback.
Thompson’s tracker can continue to serve as a list of proposed milestones, but its status changes should be read alongside the underlying sources and the tracker’s own definitions. The review does not identify a forthcoming independent assessment or a set date for resolving the disputed claims. Until those details emerge, the “early ASI” label remains an interpretation, while the specific mathematical and technical results can be judged individually.
Key Questions
What did Thorsten Meyer report?
Meyer published a source-check of 18 claims drawn from Alan D. Thompson’s 50-sign ASI tracker. He classified seven as verified, five as vendor claims, three as single-source claims and three as disputed or contradicted.
How many signs does Thompson’s tracker mark as achieved?
The tracker’s status grid, last updated on Oct. 26, 2026, lists two signs as achieved, 33 as partially achieved and 15 as not achieved. Thompson’s partial category can include enabling work and funded programmes.
Does the review establish that current models are ASI?
No. The “early ASI” description is Thompson’s assessment. Meyer’s review says definitions differ and the claims do not establish that current systems meet a shared, independently validated ASI standard.
What kinds of AI progress does the review find best supported?
Meyer says the most checkable examples cluster around mathematical results, a validated physics calculation and regulatory filings. Even the mathematics examples include caveats about human feedback, coaching or the scope of the problem.
Source: ThorstenMeyerAI.com
