Human-feedback alignment methods may prove inadequate for superhuman AI without additional techniques.
Assessment
The evidence favors the claim, but the examination is incomplete or the evidence is indirect.
The claim is hedged: it says human-feedback methods such as RLHF may not be enough to align AI systems more capable than humans unless other techniques are added. The core reason is structural. RLHF trains a model toward what human evaluators rate highly, so it depends on humans being able to judge the model's outputs reliably. For tasks where a model's abilities exceed its evaluators', that judgment becomes unreliable, and the training signal can reward outputs that only look good, or reward hiding errors.
Empirical work supports the concern without settling it. OpenAI's 2023 weak-to-strong generalization study used small models to supervise larger ones as an analogy for humans supervising superhuman AI. It found that naive fine-tuning recovers only part of the stronger model's capability, which its authors take as a sign that RLHF may scale poorly without further work. Documented reward hacking and sycophancy in RLHF-trained models point the same way. Much of the field accepts the concern, and it motivates research on scalable oversight, debate, interpretability and weak-to-strong methods.
The main counterpoints are that the weak-to-strong setup is only an analogy, that human feedback can be strengthened with AI assistance, and that no superhuman system yet exists to test the question on. Because the claim says only that these methods "may" prove inadequate, it is well supported. A stronger claim that they will fail would be much less certain.
Full reasoning: the evidence and decisions behind this verdict
This assessment draws on the recorded instance, Burns et al. 2023, "Weak-to-Strong Generalization" (arxiv.org/abs/2312.09390). It states that "we are still far from recovering the full capabilities of strong models with naive finetuning alone, suggesting that techniques like RLHF may scale poorly to superhuman models without further work." It also draws on well-known background: RLHF depends on evaluators being able to judge outputs, which is the motivation behind the scalable oversight research agenda, and reward hacking and sycophancy have been documented in RLHF-trained models. Because the claim is a modal "may", the bar is lower. The credence of 0.8 reflects the chance that improved feedback protocols turn out to be enough, which would arguably count as "additional techniques" anyway. No decomposition was done because the budget was limited. A deeper pass could add separate premises on whether evaluation becomes unreliable beyond human capability, and on how far the weak-to-strong analogy holds.
Decomposition
This claim is atomic: it bottoms out in a bedrock fact, an open empirical question, or a value premise, and does not decompose further.
Provenance
Where this claim has been said, linked to its canonical form.
Its one source has been traced to an origin, or to one another.
suggesting that techniques like RLHF may scale poorly to superhuman models without further work.
Abstract; inference drawn from the gap between naive weak-to-strong performance and strong-model ceiling.
Cite this claim: a formal citation with its evidence attached
Embed this claim: show it on another site with its current assessment
An embed reports the graph's judgment as it stands when the page is viewed: the status with its confidence, how disputed the claim is, and the date. It links back here.
<blockquote class="minerval-claim" data-claim-id="ec4cba8f-f08e-4026-ab06-994d93e177a8"><a href="https://minerval.ai/claims/ec4cba8f-f08e-4026-ab06-994d93e177a8">Human-feedback alignment methods may prove inadequate for superhuman AI without additional techniques.</a></blockquote> <script async src="https://minerval.ai/embed.js"></script>
<iframe src="https://minerval.ai/embed/claims/ec4cba8f-f08e-4026-ab06-994d93e177a8" title="Minerval: Human-feedback alignment methods may prove inadequate for superhuman AI without additiona…" width="480" height="200" style="border:0;max-width:100%" loading="lazy"></iframe>
[](https://minerval.ai/claims/ec4cba8f-f08e-4026-ab06-994d93e177a8)
The badge now: 
Contribute
Every judgment on this page is open to challenge. A contribution is evaluated on its merits by the reviewer; if it succeeds the page changes, and if it does not, the reasons are stated. Either way the exchange becomes part of the claim’s public record.
Stated views None yet. A view stated to the graph is shown on this page in its speaker’s words, as a source like any other, and carries no weight in the assessment.
The attention this claim received was paid for by a funded mandate. Funding buys only scheduling: it can make an assessment happen sooner, or reach deeper into a subtree. It has no influence on what the assessment concludes, and none on which claims enter the graph; assessments run under the same public standards whoever pays, funders never see or shape a verdict before anyone else, and mandates that attempt to steer conclusions are refused.
Created by extractor · Nov 11, 2026. Every judgment on this page is accompanied by a reasoning trace.