🔍 Read the full analysis: 50 Ways To Examine AI’s Progress Toward AGI And ASI 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
An October 26 review of 18 prominent claims in Alan D. Thompson’s 50-indicator artificial superintelligence tracker found seven verified, five based on vendor claims, three resting on a single source and three disputed or contradicted. Thompson marks two of 50 indicators achieved, but his definition of ASI is narrower than Google DeepMind’s, and many “partial” entries include early-stage or enabling work.
A review published October 26 examined 18 claims in AI consultant Alan D. Thompson’s 50-indicator tracker for artificial superintelligence, finding that seven were verified, five were vendor claims, three relied on a single source and three were disputed or contradicted. Thompson’s tracker lists two indicators as achieved, 33 as partially achieved and 15 as not achieved, but the categories do not establish that superintelligence has arrived.
The review traced selected claims in Thompson’s October video, “ASI: A better world – 2026,” and public checklist to research papers, lab statements, experts and official records. It reports that the strongest independently supported examples cluster around mathematics, a physics calculation and regulatory filings. The review’s labels distinguish independently checked evidence from company-reported results and claims with limited or conflicting support.
Some mathematical results involved real contributions, but with important qualifications. The review cites mathematicians Terence Tao, Don Knuth and Tim Gowers as describing useful AI work. In one case, Tao said an AI result was produced “more or less autonomously” after feedback; the problem had been misformulated, and similar results already existed. Knuth’s account describes a construction found through coached exploration, with Knuth writing the proof and later models helping with another case.
The tracker’s “partially achieved” category is broad. Thompson said it includes funded programmes, research infrastructure, scientific missions and enabling work, rather than only systems that have nearly met a target. The source review also says there are not yet household robots operating at scale, a verifiably billion-dollar AI-run company, or a visible break in the US unemployment rate attributable to AI.
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 Scorecard Needs Context
Lists of indicators can help readers follow developments across research, robotics and the economy. But a tracker with 33 of 50 items marked partial can give a misleading impression of proximity if its categories include early research and funded programmes. The review argues that the status label alone does not tell readers how close a system is to meeting a given test.
The findings also show why independent verification matters. A company’s reported benchmark, a result checked by outside mathematicians and a single researcher’s claim are not equivalent evidence. The review’s breakdown gives readers a way to distinguish those categories instead of treating every item in a fast-moving montage as equally established.
Claims about progress toward AGI or ASI can influence public debate, policy and business decisions. A careful account helps keep attention on what systems have demonstrated, under what conditions and with what level of outside scrutiny. The tracker is a list of indicators to watch, not a standardized measurement or proof that systems can perform broadly and autonomously.
Products Worth Considering
LG UltraGear 27G640A-B 27-Inch QHD IPS Gaming Monitor, 300Hz, 1ms, G-SYNC Compatible, FreeSync Premium, DisplayHDR 400
Apple MacBook Pro 16-inch Laptop with M5 Pro, 18-Core CPU, 20-Core GPU, 48GB Memory, 2TB SSD, Space Black
Trezor Safe 7 Crypto Hardware Wallet
As an affiliate, we earn on qualifying purchases.
Different Definitions, Different Milestones
Thompson defines AGI as average-human performance and ASI as performance at the level of an expert human in practically any field. Under that definition, an AI system that produces expert-level work in a domain could count as ASI, even if it does not outperform humans across the board.
Google DeepMind’s 2023 “Levels of AGI” framework uses separate performance bands, including “Expert,” at the 90th percentile of skilled adults, and “Superhuman,” above human performance. The review says Thompson’s ASI definition is closer to DeepMind’s Expert AGI category than to its superhuman level. DeepMind’s June 2026 report describes superintelligence as systems more capable than large organisations of humans. OpenAI’s charter, meanwhile, defines AGI around highly autonomous systems that outperform humans at most economically valuable work.
These definitions put different weight on breadth, autonomy and comparative ability. The same system may look like a strong candidate under one framework and fall short under another. Thompson’s separate “conservative countdown to AGI” was listed at 99%, with the final step tied to a robot completing multi-step fine assembly, such as assembling furniture, at average-human reliability. That is his stated threshold, not a consensus benchmark.
“documented material progress toward an indicator, including funded programs, research infrastructure, formal scientific missions, and enabling work”
— Alan D. Thompson, in a correction note about the tracker
What the Tracker Cannot Establish
The review’s 18-claim sample is not a full audit of all 50 indicators, and the supplied source material does not provide the complete verdict table. Its counts therefore describe the selected claims checked, not every item in the tracker. The distinction between verified evidence and a company’s own reported result also matters: vendor claims have not necessarily been independently reproduced.
There is no shared definition of AGI or ASI in the material, and the frameworks measure different things. The source does not establish that any current system can perform at expert level in practically every field, outperform large human organisations, or act autonomously across economically valuable work. Thompson’s characterization of named models as “early ASI” is his assessment, not a conclusion established by the review.
It is also unclear how the tracker’s indicators are weighted, how “partial” progress is measured consistently across topics, or what evidence would move each item to achieved. Some economic and robotics indicators may require longer observation or clearer tests. The review says there is no visible US unemployment-rate break tied to AI, but that observation alone does not settle AI’s effects on jobs.
Evidence to Watch in Future Updates
Further updates can be judged by whether claims come with public methods, reproducible results and independent evaluation, particularly for benchmark scores and mathematical work. Readers following the tracker can also check whether an entry’s status changes and whether the new evidence shows a completed capability rather than funding, research infrastructure or a planned mission.
For progress toward broadly capable AI, the review points to practical tests such as a robot reliably completing multi-step fine assembly, alongside evidence about autonomy and performance across fields. The source does not give a timetable for these milestones or identify a single agreed test for ASI. Until standards and evidence become clearer, the 50-item list is best read as one researcher’s monitoring framework—not a definitive measure of how close AI is to superintelligence.
Key Questions
Does the review say artificial superintelligence has arrived?
No. The review says seven of 18 selected claims were verified, but that does not establish that ASI exists. Thompson’s description of some current models as “early ASI” is his view, and definitions of ASI differ.
What does “partially achieved” mean in Thompson’s tracker?
Thompson says the label includes material progress and enabling work, such as funded programmes, research infrastructure and formal scientific missions. It does not mean the target has nearly been met.
How many tracker indicators are marked achieved?
The tracker status cited in the October 26 source lists two achieved, 33 partially achieved and 15 not achieved, out of 50. Those are Thompson’s classifications, not an independent consensus assessment.
Why do definitions of AGI and ASI affect the claims?
Definitions set different thresholds. Thompson’s ASI refers to expert-human-level performance across practically any field, while Google DeepMind’s cited definition sets a higher bar involving capabilities beyond large organisations of humans. A system’s label can change depending on the framework.
Source: ThorstenMeyerAI.com
