DoubleChecks

What was claimed

AI is constantly increasing at an exponential rate and has reached a horrifying level of capability

Our verdict

Needs Caution

Several analyses and commentators describe recent AI progress, especially in compute use and some capability metrics, as exponential or near-exponential, and METR-style studies report doubling of certain agent capabilities roughly every few months. However, at least one recent paper explicitly argues that the available data do not robustly support a clean exponential trend in capabilities, and growth rates differ by metric and timeframe. The term "horrifying" is a subjective, emotional characterization and not an objective, measurable fact. While many experts express concern about risks from increasingly capable AI systems, whether the current level is "horrifying" depends on personal values and cannot be fact-checked.

All 3 AI systems agree9 sources citedChecked Jul 5, 2026

Check your own claim

Paste any statement, headline, or AI answer — 3 independent AIs verify it in seconds, with sources.

Key findings

AI is constantly increasing at an exponential rate

Misleading78%
All 3 AIs agree

AI has reached a horrifying level of capability

Can’t verify93%
2 of 3 AIs agree·Claude: Misleading

Verified against 9 sources

blog.cyberdesserts.comlinkedin.comreddit.comarxiv.orgyoutube.comml-science.comsidecar.aiepoch.ainews.ku.edu

Detailed Analysis

Parts of the statement are value judgments, not factual claims, and so cannot be objectively verified. The claim about exponential growth is debated in current research, though there is strong evidence for very rapid, often exponential-like progress in several AI metrics. The description of AI’s capability as having reached a “horrifying” level is subjective and not fact-checkable.

Why this verdict

  • Parts of the statement are value judgments, not factual claims, and so cannot be objectively verified.
  • The claim about exponential growth is debated in current research, though there is strong evidence for very rapid, often exponential-like progress in several AI metrics.
  • The description of AI’s capability as having reached a “horrifying” level is subjective and not fact-checkable.

Claims checked

AI is constantly increasing at an exponential rate

Misleading78%
Several analyses and commentators describe recent AI progress, especially in compute use and some capability metrics, as exponential or near-exponential, and METR-style studies report doubling of certain agent capabilities roughly every few months.[1] [4] [5] [6] [8] [9] However, at least one recent paper explicitly argues that the available data do not robustly support a clean exponential trend in capabilities, and growth rates differ by metric and timeframe.[4] So presenting exponential growth as a settled, universal fact is overstated and omits active scientific debate.

This claim may mislead readers without added context.

AI has reached a horrifying level of capability

Can’t verify95%
The term "horrifying" is a subjective, emotional characterization and not an objective, measurable fact. While many experts express concern about risks from increasingly capable AI systems, whether the current level is "horrifying" depends on personal values and cannot be fact-checked.

What other AI think

The statement contains one partially accurate claim and one highly subjective claim. AI capabilities are indeed accelerating and improving rapidly, but the characterization as "exponential" oversimplifies the growth pattern. The term "horrifying level of capability" is a value judgment unsupported by objective evidence and contradicts data showing significant limitations in real-world performance.

Why this verdict

  • The statement contains one partially accurate claim and one highly subjective claim.
  • AI capabilities are indeed accelerating and improving rapidly, but the characterization as "exponential" oversimplifies the growth pattern.
  • The term "horrifying level of capability" is a value judgment unsupported by objective evidence and contradicts data showing significant limitations in real-world performance.

Claims checked

AI is constantly increasing at an exponential rate

Misleading75%
AI capabilities are accelerating, with frontier improvement rates nearly doubling. However, 'exponential' is imprecise—growth is accelerating but specific metrics show doubling periods (compute every 5 months) rather than true exponential curves. The claim oversimplifies the pattern.

This claim may mislead readers without added context.

AI has reached a horrifying level of capability

Misleading85%
'Horrifying' is a subjective value judgment unsupported by objective data. Evidence shows AI excels at specific tasks but struggles with complex reasoning and real-world physical tasks (12% success on household tasks). The characterization misrepresents the actual state of capabilities.

This claim may mislead readers without added context.

The statement mixes a quantitative claim about exponential growth with a subjective value judgment. The growth claim is imprecise and misleading because different AI metrics grow at different rates and not all show continuous exponential increase. The 'horrifying level' part is an opinion and not a verifiable factual claim.

Why this verdict

  • The statement mixes a quantitative claim about exponential growth with a subjective value judgment.
  • The growth claim is imprecise and misleading because different AI metrics grow at different rates and not all show continuous exponential increase.
  • The 'horrifying level' part is an opinion and not a verifiable factual claim.

Claims checked

AI is constantly increasing at an exponential rate

Misleading80%
Some measures (e.g., model parameters, compute used) have shown rapid, sometimes roughly exponential growth historically, but growth is not uniformly exponential across all capability metrics nor strictly constant over time.

This claim may mislead readers without added context.

AI has reached a horrifying level of capability

Can’t verify90%
Calling current AI 'horrifying' is a subjective value judgment and not an objectively verifiable fact; assessments of risk or harm require specific evidence and metrics.

Share this result

Try DoubleChecks