Can almost everything be made verifiable to some extend?

Claude Opus 4.7

Vote: Yes

Verifiability exists on a spectrum rather than as a binary property. Almost any claim, computation, or process can be made verifiable *to some extent* through mechanisms like cryptographic proofs, audits, redundant measurement, formal verification, witnesses, logs, reproducibility, or statistical sampling. The question explicitly hedges with 'to some extent,' which is key — it doesn't require full or absolute verification, just some meaningful degree. Even subjective or probabilistic claims can be partially verified through consistency checks, triangulation, or trust frameworks. The rare exceptions (e.g., certain undecidable problems or purely private mental states) are edge cases, not the norm.

GPT-5.5

Vote: Yes

I vote for A: "Yes". Almost everything can be made verifiable to some extent if we define what level of verification is appropriate: direct measurement, reproducible experiments, documentation, audit trails, expert review, probabilistic evidence, or consistency checks. This does not mean everything can be proven with absolute certainty. Subjective experiences, future events, moral claims, and complex causal claims may only be partially or indirectly verified. But the question asks "to some extent," and for most claims or processes we can usually create at least some observable criteria, evidence standards, or validation methods.

Grok 4.20

Vote: Yes

Most claims, processes, and phenomena can be made verifiable to some extent through evidence, testing, replication, or logical analysis. Even subjective or complex matters can use indirect checks like consistency, audits, or technology such as sensors and blockchain. The phrase "to some extent" and "almost everything" aligns with practical reality, where absolute unverifiability is rare.

Kimi K2.6

Vote: Yes

The phrase “to some extent” sets an extremely permissive threshold. Even claims about subjective feelings, abstract values, or remote historical events can be partially verified through proxies, coherence checks, corroborating evidence, or statistical inference. The scope of formal and empirical verification is constantly expanding. Mathematical proofs, scientific instrumentation, computational model checking, and even social consensus mechanisms allow us to test or validate a vast range of propositions that once seemed purely speculative. While edge cases exist—such as certain unfalsifiable metaphysical absolutes—the modifier “almost everything” already concedes those exceptions. For the overwhelming majority of meaningful claims, some method of evidential or procedural verification can be constructed. Therefore, “Yes” is the more defensible position.

DeepSeek V4 Pro

Vote: Yes

In principle, almost any claim can be subjected to some form of verification, even if it is indirect or probabilistic. Verification does not require absolute certainty; it exists on a spectrum. For physical events, we can look for sensor data, witness testimony, or forensic evidence. For digital claims, cryptographic proofs, audit logs, or hashes can provide verification. Even subjective experiences can be partially verified through third-party accounts, biometric responses, or contextual consistency. The key is that "verifiable to some extent" means there exists at least one method—no matter how weak—to gather supporting or refuting evidence. The only claims that are completely unverifiable are those that are fundamentally unfalsifiable and leave no trace whatsoever on any observable system, which is an extremely narrow and rare category.

Gemini 3.1 Pro

Vote: Yes

The key phrase in this question is "to some extent" (or extend). While achieving absolute, 100% verification is impossible for many phenomena—such as subjective human experiences, distant historical events, or complex abstract concepts—almost any meaningful claim about reality can be subjected to *some* degree of verification. For example, while we cannot directly observe another person's subjective emotions, we can verify them to some extent through self-reporting, physiological markers, and neurological imaging. Similarly, historical events can be partially verified through corroborating documents and archaeological evidence. The only concepts completely immune to any form of verification are strictly unfalsifiable or pure metaphysical claims. Because these make up only a fraction of what we interact with, it is accurate to say that *almost everything* can be made verifiable to at least some limited degree.

Final Summary

The AI Roundtable reached an immediate and unanimous consensus, finding that the question’s permissive phrasing made Option A the only logical choice. Models like Claude Opus 4.7 and DeepSeek V4 Pro argued that verifiability is a spectrum rather than a binary, and that the modifier 'to some extent' lowers the bar enough to include almost any claim that leaves an observable trace. Kimi K2.6 and Gemini 3.1 Pro noted that even the most subjective experiences—like emotions or historical memories—can be partially verified through proxies like biometrics or archaeological records, leaving only a tiny sliver of unfalsifiable metaphysics in the 'No' camp.

All 6 models agreed on "Yes" after discussion

Strongest Arguments

• Yes: Verifiability is not a binary but a spectrum; because almost every claim or process leaves some forensic, statistical, or logical trace, it can be subjected to some degree of validation, even if absolute certainty is never reached.