🔍 Read the full analysis: Where Does AI Stand On The Path To AGI And ASI? 50 Signs Checked on ThorstenMeyerAI.com
Get the latest gadgets 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 researcher Alan D. Thompson’s 50-sign ASI tracker found seven verified, five based on vendor claims, three supported by a single source and three disputed or contradicted. Thompson’s tracker labels two signs achieved and 33 partially achieved, but its definition of ASI is closer to expert-level performance than some other research frameworks’ definition of superintelligence.
A review published October 26 of 18 prominent claims in researcher Alan D. Thompson’s 50-sign artificial superintelligence tracker found seven supported by independent experts or official records, five resting on companies’ own claims, three based on a single source and three disputed or contradicted. Thompson’s tracker marks two of its 50 indicators achieved and 33 partially achieved, but those labels do not establish that AI systems have reached superintelligence.
Thompson, an Australian AI consultant who publishes The Memo newsletter, lists indicators ranging from mathematical discoveries to household robots and space exploration. His tracker, last updated October 26, classifies 15 indicators as not achieved. In an August correction note, he explained that “partially achieved” covers documented progress that can include funded programmes, research infrastructure and enabling work. The category therefore does not necessarily mean a system is close to meeting an indicator.
The review examined 18 headline claims from Thompson’s 2026 video montage and tracker, checking primary sources such as papers, researchers’ statements, company posts and regulatory filings. It says the stronger independently checked examples tend to involve results that can be inspected, including mathematical work, a physics calculation and regulatory documents. The review also reports that mathematicians Terence Tao, Don Knuth and Tim Gowers have described real AI contributions, while noting qualifications around human guidance, existing literature and the kind of problem being solved.
Not all claims carry the same evidentiary weight. For example, the review labels OpenAI’s account of a model producing results on decade-old mathematical problems for about $2,000 a vendor claim, while noting that Lean certificates were made public and that human work was involved. It describes other reported mathematical results as independently verified, but the available source material does not provide the full evidence or assessment for every one of the 18 claims.
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 Scrutiny
The tracker offers readers a structured way to follow developments, but its totals can sound more definitive than the underlying categories warrant. With 33 indicators marked partial, the result depends heavily on what qualifies as progress. Because that status can include funded programmes and supporting research, it should not be read as 33 signs that systems are nearly superintelligent.
Definitions also change the conclusion. Thompson calls a system ASI if it performs at an expert human level in practically any field. Google DeepMind’s 2023 framework separates “Expert” performance from “Superhuman,” while its June 2026 report describes superintelligence as exceeding the cognitive capabilities of large organisations of people. A model may therefore count as ASI under Thompson’s definition without meeting another framework’s threshold.
For readers, the distinction matters when assessing claims about how close AI is to broad human-level capability, economic change or independent scientific work. Verified results in narrow domains are meaningful developments, but they do not by themselves establish general, autonomous competence across fields.
Products Worth Considering
Apple iPad Air (2025) 13‑inch Wi‑Fi 256 GB – Space Gray (M3 chip)
LG UltraGear 32GX850A-B 32-Inch 4K UHD Glossy OLED Gaming Monitor
Logitech MX Mechanical Wireless Illuminated Keyboard, Tactile Quiet, Graphite
As an affiliate, we earn on qualifying purchases.
Different Benchmarks, Different Thresholds
Thompson defines AGI as a machine performing at the level of the median human and ASI as one performing at the level of an expert human in practically any field. Those are his stated thresholds, not a universal standard. Google DeepMind’s 2023 “Levels of AGI” framework labels the 90th percentile of skilled adults “Expert,” then distinguishes “Virtuoso” and “Superhuman” performance above that. Its June 2026 report uses a more expansive description of superintelligence, based on capability beyond large human organisations.
OpenAI’s charter, as cited in the source material, defines AGI in terms of highly autonomous systems that outperform humans at most economically valuable work. That emphasis on autonomy and breadth differs from a scorecard that can count a strong result in a particular field. Thompson also maintains a separate “conservative countdown to AGI,” reported at 99%; he says the remaining hurdle is a robot completing multi-step fine assembly, such as assembling furniture, at average-human reliability. That figure is his estimate, not an independently validated measurement of AGI progress.
“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
What the Evidence Cannot Settle
The review’s classifications do not amount to a single agreed measurement of AGI or ASI. Its 18 selected claims are a sample of the tracker’s claims, not an assessment of all 50 indicators, and the source material does not provide the full documentation for every verdict. A vendor claim can be informative, but it is not equivalent to independent reproduction; a single-source report carries different limitations from an official filing or an expert-checked result.
It also remains unclear how much current AI capability can be transferred from specific, often tool-assisted demonstrations to reliably autonomous work across unrelated fields. The source says some mathematical achievements involved coaching or human proof-writing, and that several other claims are disputed or contradicted. It reports no household robots deployed at scale, no verifiably billion-dollar AI-run company, and no visible break in the US unemployment rate attributable to AI. These observations do not resolve the broader economic or technical questions.
Evidence to Watch in Future Updates
The next useful test is whether new claims come with public methods, reproducible results and independent evaluation, rather than a status label alone. Readers can also watch for evidence that systems complete extended tasks with less human guidance and perform reliably beyond the domain in which they were tested.
Thompson’s tracker is a continuing list, and the review describes its status grid as updated on October 26, 2026. Future changes may add evidence or alter classifications; the source material does not specify a scheduled next update. Until stronger cross-domain and autonomy evidence is available, the tracker is best read as one researcher’s monitoring framework, not a definitive measure of how close AI is to AGI or ASI.
Key Questions
What did the review find about the 18 ASI claims?
It classified seven as verified, five as vendor claims, three as based on a single source and three as disputed or contradicted. The review’s counts cover the 18 claims it selected, not every item in the tracker.
Does “partially achieved” mean AI is nearly superintelligent?
No. Thompson says the label includes supporting progress such as funded programmes and research infrastructure. It does not necessarily mean the system has met, or is close to meeting, the indicator.
Why do researchers disagree about whether AI has reached ASI?
They use different definitions. Thompson’s ASI threshold is expert-level performance in practically any field, while other frameworks put greater emphasis on superhuman capability, broad autonomy or performance beyond large human organisations.
What evidence does the source say is still missing?
The source reports no household robots at scale and says the final hurdle in Thompson’s separate AGI countdown is reliable, multi-step fine assembly by a robot. It also says there is no visible break in US unemployment attributable to AI, while that observation does not settle the wider economic impact.
Source: ThorstenMeyerAI.com
![Stepping Inside The Design: How VR Is Advancing USACE Projects [Image 3 Of 7]](https://techieus.com/wp-content/uploads/2026/08/stepping-inside-the-design-how-vr-is-advancing-usace-projects-image-3-of-7-featured-560x300.jpg)
