🔍 Read the full analysis: 50 Signs, One Question: How Close Is AI To AGI And ASI? 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
A review published October 26 examined 18 prominent claims from researcher Alan D. Thompson’s 50-indicator tracker for artificial superintelligence. Seven claims were verified, five were vendor claims, three relied on a single source and three were disputed or contradicted; the tracker itself marks only two indicators achieved. The findings show real progress in some research tasks, but do not establish that AI has reached AGI or ASI.
A review published October 26 examined 18 prominent claims from Australian AI consultant Alan D. Thompson’s 50-indicator artificial superintelligence tracker, finding seven verified, five based on company claims, three supported by a single source and three disputed or contradicted. Thompson’s tracker says two indicators are achieved, 33 partially achieved and 15 not achieved, a tally that does not establish that current systems are AGI or ASI.
The tracker ranges from claims about mathematical research and self-improving code to far-future scenarios. Thompson describes current systems including Claude Mythos Preview and GPT-6 Astra as “early ASI.” That is his assessment, not a shared industry measurement. His definition of ASI is a machine performing at expert-human level in practically any field.
In the review’s sample of 18 headline claims, evidence was strongest where results could be checked against research, expert accounts or official records. The reviewed examples included AI contributions to mathematical problems and a claim about a conjecture that was examined by external mathematicians. Other entries relied on company announcements or had limited independent support. The source review cautions that a public formal proof or certificate can support a result without independently confirming every claim about how it was produced.
“Partially achieved” is a broad category in Thompson’s tracker. He said in an August correction note that it covers “documented material progress toward an indicator, including funded programs, research infrastructure, formal scientific missions, and enabling work.” That means orange status can reflect groundwork rather than a capability close to completion. The tracker, for example, places two exoplanet candidates identified with Claude Code under an indicator about colonizing another planet—a classification that should not be read as evidence that colonization has occurred.
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 Varies by AI Capability
The review matters because claims that AI is nearing superintelligence can combine demonstrated results, company-reported performance and speculative milestones. Treating all of them as equivalent can give readers a misleading impression of how close systems are to operating reliably across domains or acting independently.
The clearest signs in the reviewed material concern bounded tasks, especially mathematics and scientific work where experts can inspect the output. That does not settle whether AI can perform the full range of work associated with AGI, much less do so autonomously and more capably than large human organizations. Those distinctions affect public debate about labor, research, safety and the pace of deployment.
The tracker also reports no household robots operating at scale, no verifiably billion-dollar AI-run company and no visible break in the US unemployment rate, according to the source review. These observations are not a complete test of AGI, but they underline the gap between progress on selected benchmarks or research problems and broad economic or physical-world capability.
Products Worth Considering
Apple iPad Air (4th Generation, 10.9-inch, Wi-Fi, 256GB) – Space Gray
LG 34GX90SA-W 34-inch Ultragear WQHD OLED Curved Gaming Monitor
Bose TV Speaker (2nd Gen) Compact Soundbar with Bluetooth and HDMI Connectivity
As an affiliate, we earn on qualifying purchases.
Different Scales, Different Answers
Thompson defines AGI as a machine performing at the level of an average human and ASI as one performing at expert-human level in practically any field. Google DeepMind’s 2023 “Levels of AGI” framework uses a different ladder, ranging from “Competent” through “Expert” and “Virtuoso” to “Superhuman.” Under that scale, Thompson’s ASI threshold roughly corresponds to “Expert” performance, not the framework’s superhuman level.
DeepMind’s June 2026 report, “From AGI to ASI,” describes superintelligence as a system more intelligent and cognitively capable than large organizations of humans. OpenAI’s charter, meanwhile, defines AGI around highly autonomous systems that outperform humans at most economically valuable work. These definitions place different weight on breadth, autonomy and performance relative to people, so the same model can appear closer to one threshold than another.
Thompson also maintains a separate “conservative countdown to AGI,” which the source says stood at 99%. He identifies the remaining requirement as a robot completing multi-step fine assembly, such as putting together flat-pack furniture, with average-human reliability. That figure is his judgment about a selected milestone, not an independently established probability that AGI is near.
“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 his tracker
What the Tracker Cannot Establish
The review is a check of 18 selected claims, not an audit of all 50 indicators or a consensus assessment by the AI research field. Its verdict categories also differ in strength: a verified result in a mathematical task does not verify that a system is generally intelligent, while a vendor claim may be accurate but lacks independent reproduction in the review.
It remains unclear how the reviewed systems would perform across the full range of tasks implied by the competing definitions, and how reliably they could act without human direction. The supplied material does not provide enough information to independently assess every claim in the montage, including the complete methods behind company-reported results. Thompson’s label “early ASI” should therefore be understood as his interpretation, not a confirmed scientific status.
Tests That Could Change the Picture
The next meaningful evidence will come from independent replication, clearer reporting of methods and performance, and demonstrations that extend beyond narrow or coached research tasks. For claims about robotics and autonomy, readers will need evidence of dependable performance in varied real-world settings, not just progress toward a milestone.
Thompson’s tracker may change as new results emerge, but its colors should be read according to his stated criteria. Until researchers and labs use more comparable definitions and publish results that can be independently checked, the question of how close AI is to AGI or ASI has no single measured answer.
Key Questions
Does the review conclude that AI has reached AGI or ASI?
No. The review found mixed evidence across 18 selected claims and does not establish that current systems meet a general AGI or ASI threshold.
What does Thompson’s tracker report?
As of its October 26 update, it marks two indicators achieved, 33 partially achieved and 15 not achieved. Thompson’s partial category can include funded programs and enabling work.
Why do definitions of ASI matter?
Different frameworks set different bars. Thompson describes expert-level performance across fields as ASI, while DeepMind’s 2026 description compares a system with the cognitive capabilities of large organizations of humans.
What evidence was strongest in the claims reviewed?
The source review says stronger evidence tended to involve results that experts or official records could check, including selected mathematical contributions. Other claims were based on company reporting or limited sources.
Source: ThorstenMeyerAI.com
