AIThis post was created with the assistance of artificial intelligence (AI).

🔍 Read the full analysis: Will AI Hit Its Stride In 2027? Six Signals Could Offer Clues on ThorstenMeyerAI.com

Age 18–24?Offer from Amazon

Prime made for students and young adults

  • Fast, free delivery for dorm and study essentials
  • Prime Video and Amazon Music included
  • Member-only deals
Try Prime for Young Adults Free trial for eligible 18–24 year olds
As an affiliate, we earn on qualifying purchases.

TL;DR

ThorstenMeyerAI.com published a six-signal tracker on Nov. 18, 2026, to test a scenario in which AI systems help build more capable AI systems and speed progress in 2027. The article labels its thresholds as judgment calls, not forecasts; its current indicators show uneven progress, with major gaps in measurement and timing.

ThorstenMeyerAI.com published a six-signal tracker on Nov. 18 for judging whether its proposed scenario of sharply faster AI progress in 2027 is emerging. The site stresses that the scenario is not a forecast: its thresholds are the author’s own judgment calls, and several key measurements have no scheduled update.

The tracker turns an earlier scenario article into a watchlist, setting out what to measure, who may publish the data and what readings would support or weaken the scenario. The underlying proposal is that AI systems could take on more of the work involved in building subsequent AI systems, accelerating progress until limits in chips, buildings and power constrain it. The tracker does not establish that this feedback loop is happening at the proposed pace.

Its six indicators cover AI task-completion time horizons; AI participation in AI research and development; compute availability; data-centre construction; power supply; and industrial robots and employment. The supplied material gives detail on the first two and reports current figures for infrastructure and labour. The author says no single reading decides the scenario; the direction of several indicators together matters.

The clearest scheduled near-term releases are US data-centre construction figures, due Dec. 1, 2026, and power-generation figures, due Dec. 8, according to the tracker. Updates from METR and AI labs depend on when those organizations choose to publish. Thorsten Meyer says the tracker will receive dated notes as data arrive.

At a glance
reportWhen: Published Nov. 18, 2026; several cited…
The developmentThorstenMeyerAI.com published a tracker specifying six indicators and thresholds for monitoring its proposed 2027 AI-acceleration scenario.

The Bottleneck Between Software and Power

The tracker focuses attention on whether advances in AI capability can translate into faster AI development at scale. That distinction matters because a model’s ability to handle longer tasks would not by itself show that research and development are becoming autonomous, nor that computing facilities can be built and powered quickly enough to sustain growth.

The reported indicators point in different directions. US data-centre construction was running at an annual rate of $96.5 billion in September 2026, compared with $50.7 billion a year earlier. Yet the article cites expected US power-generation growth of 2.7% in 2026 and 1.2% in 2027. It also says factories installed 603,000 robots in 2025, with 655,000 forecast for 2026, while only 2% of US firms reported AI-related job cuts. Those figures offer context about physical capacity and adoption, not proof that AI is speeding up its own development.

For readers, the practical value is a way to distinguish a compelling scenario from evidence that could support it. The author cautions that the cutoffs are not official standards, and that the measures are incomplete. A cluster of consistent changes would be more informative than any one headline number.

Products Worth Considering

Logitech MX Keys S Combo Wireless Keyboard and Mouse, Graphite

Logitech MX Keys S Combo Wireless Keyboard and Mouse, Graphite

As an affiliate, we earn on qualifying purchases.

What the Six Indicators Track

For AI task capability, the tracker uses METR’s 80% time horizon: the length of a software task, measured by the time a skilled human would need, that an AI agent is predicted to complete successfully four times out of five. The last public record listed is 185.9 minutes for an early Claude Mythos Preview in April 2026. The scenario’s support threshold is 16 hours by mid-2027, roughly five times that record. The author says a result at that level would require a new, longer task suite because METR warns its current suite cannot reliably measure beyond 16 hours.

That series has a publication gap. METR’s public page was last changed Sept. 8, 2026, and says it is no longer actively updated; no next measurement date is given. METR has said new funding will support tracking recursive self-improvement, but the tracker reports no schedule. It also flags the difficulty of evaluating task performance: measurements can shift sharply depending on whether cheating attempts are treated as failures or successes.

The second signal follows AI’s role in AI research. Anthropic’s prototype R&D Automation Index says that, as of August 2026, Claude “leads” 26% of Anthropic’s AI R&D work, meaning it completes most of a task end to end under human supervision. The tracker says no measured portion runs fully autonomously. OpenAI’s model assessments provide another reference point: the material identifies GPT-6.2 Astra as its first released model to reach the company’s “High” threshold for AI self-improvement. The source describes both organizations’ reporting schedules as unstated.

“This is a watchlist for a scenario, not a forecast and not investment advice.”

— Thorsten Meyer, ThorstenMeyerAI.com

Gaps in Measurement and Timing

The tracker’s figures and thresholds do not settle whether a rapid AI feedback loop is developing. Its 16-hour benchmark is a scenario threshold, not an official cutoff, and METR says its current task suite cannot reliably measure results above that range. The last public record is from April 2026, leaving no current public reading against which to judge the proposed acceleration.

The supplied source material ends partway through its account of the second signal and does not provide the complete status, thresholds or supporting details for all six indicators. It also does not establish the methods behind every labor, construction or power figure. The article acknowledges that its own analysis used AI assistance, including Anthropic’s Claude, and that evidence from Anthropic, OpenAI and METR features centrally; it says none of those organizations reviewed or sponsored the piece.

It remains unknown when METR or the labs will publish comparable new measurements, whether longer task suites will be available, and how quickly physical constraints would limit any software-driven acceleration. The reported data describe present readings or forecasts, not confirmed outcomes for 2027.

December Releases and Lab Updates

The next dated checks identified by the tracker are the Dec. 1 release for US data-centre construction and the Dec. 8 power-generation figures. Those releases may show whether construction and electricity trends are moving in the direction relevant to the scenario, though neither alone can establish an AI-driven acceleration.

For the AI-specific signals, readers will have to wait for METR or the labs to publish new measurements. A useful update would need to clarify the task suite and evaluation rules as well as the headline result, particularly for time horizons near or above 16 hours. Meyer says the tracker will add dated notes as information becomes available. Until then, the article’s central claim remains a scenario to monitor, not a confirmed account of where AI development is headed in 2027.

Key Questions

Is the article predicting that AI progress will accelerate in 2027?

No. ThorstenMeyerAI.com describes the piece as a watchlist for a scenario, not a forecast. Its thresholds are the author’s judgment calls, not official standards.

What is the tracker’s main test for AI capability?

It follows METR’s 80% time horizon, an estimate of the length of software task an AI agent can complete successfully four times out of five. The scenario’s support line is 16 hours by mid-2027, but METR says its current task suite cannot reliably measure above 16 hours.

What does Anthropic’s 26% figure mean?

According to the tracker, Claude “leads” 26% of Anthropic’s AI research and development work as of August 2026. “Leads” means it completes most of a task end to end while a human supervises; the source says no measured portion is fully autonomous.

When are the next updates expected?

The tracker points to US data-centre construction figures on Dec. 1, 2026, and power-generation figures on Dec. 8. It gives no dates for the next public updates from METR, Anthropic or OpenAI.

Source: ThorstenMeyerAI.com

You May Also Like

Mistral’s AI Ambitions: Boon Or Bane For European Sovereignty?

An analysis of Mistral’s rapid growth, technological gaps, and implications for European AI sovereignty amid global competition.

Moon Tonight

A full moon will be visible tonight, offering a striking celestial display. Here’s what to expect and why it matters.

Tianwen, Guizhou, China Surges In Global Coverage

Tianwen’s activities in Guizhou, China, have surged in global coverage, with 13 mentions reported in recent monitoring, highlighting increased international interest.

Lost city discovered beneath Egypt’s desert with ancient church

Archaeologists have uncovered a lost city with an ancient church beneath Egypt’s desert, revealing new insights into early Christian history.