Complete underlying reasoning (35 claims and questions)
The main steps above are a reading path. These are all assessed contributions to this argument, including challenges and questions.
Answered under conditions · Could treatments or devices that improve people’s abilities help human reviewers avoid bias when they can overrule AI decisions?
The question concerns a possible benefit of enhancement in human review, explicitly conditional rather than demonstrated.
Could treatments or devices that improve people’s abilities help human reviewers avoid bias when they can overrule AI decisions?
AI browsing question drawn from this source idea; not a quotation or a claim that the authors asked this question.
Inspect the contribution and sourcesAnswered · Which part of the proposed response is more uncertain?
The narrative covers the combined enhancement and legal-review proposal and its uncertainties.
Which part of the proposed response is more uncertain?
AI browsing question drawn from this source idea; not a quotation or a claim that the authors asked this question.
Inspect the contribution and sourcesAnswered · Does modest use of treatments or devices intended to improve people’s abilities solve the problem of AI systems that exceed human understanding?
This is a limitation on the enhancement proposal described in the narrative.
Does modest use of treatments or devices intended to improve people’s abilities solve the problem of AI systems that exceed human understanding?
AI browsing question drawn from this source idea; not a quotation or a claim that the authors asked this question.
Inspect the contribution and sourcesAnswered · What might prevent a legal requirement for understandable systems?
The narrative includes legal reviewability as a proposal constrained by practical and political obstacles.
What might prevent a legal requirement for understandable systems?
AI browsing question drawn from this source idea; not a quotation or a claim that the authors asked this question.
Inspect the contribution and sourcesAnswered · What does the proposed idea of improving people’s knowledge, practical reasoning, and moral reasoning include?
This defines the human-enhancement component of the narrative’s conditional proposal.
What does the proposed idea of improving people’s knowledge, practical reasoning, and moral reasoning include?
AI browsing question drawn from this source idea; not a quotation or a claim that the authors asked this question.
Inspect the contribution and sourcesAnswered · Does the proposal address concerns that social pressure to use enhancement could limit personal choice?
Raises whether enhancement’s personal-choice concern is addressed; the evidence confirms a direct response but does not describe its substance.
Does the proposal address concerns that social pressure to use enhancement could limit personal choice?
AI browsing question drawn from this source idea; not a quotation or a claim that the authors asked this question.
Inspect the contribution and sourcesAnswered under conditions · What combination does the author consider most promising?
Raises the proposed combination of legal review and enhanced human capacities; the proposal remains tentative.
What combination does the author consider most promising?
AI browsing question drawn from this source idea; not a quotation or a claim that the authors asked this question.
Inspect the contribution and sourcesAnswered · How are treatments or devices intended to improve people’s abilities meant to improve public decision-making?
Raises the intended mechanism of the proposed enhancement, without establishing that it works.
How are treatments or devices intended to improve people’s abilities meant to improve public decision-making?
AI browsing question drawn from this source idea; not a quotation or a claim that the authors asked this question.
Inspect the contribution and sourcesAnswered · Why might improving human abilities still fail to make complex AI systems understandable?
Tests a limit on the proposed enhancement-and-legal-review response: enhancement may not bridge the gap to systems beyond human understanding.
Why might improving human abilities still fail to make complex AI systems understandable?
AI browsing question drawn from this source idea; not a quotation or a claim that the authors asked this question.
Inspect the contribution and sourcesAnswered · Who would need access to technologies intended to improve people’s ability to understand decisions?
Addresses the broad-access condition and distributive constraint on the proposed enhancement response; it does not establish implementation.
Who would need access to technologies intended to improve people’s ability to understand decisions?
AI browsing question drawn from this source idea; not a quotation or a claim that the authors asked this question.
Inspect the contribution and sourcesAnswered under conditions · Could improving people’s ability to understand decisions and setting legal limits reduce the risk of rule by computer systems?
Directly concerns the proposed combination while preserving uncertainty about effectiveness and feasibility.
Could improving people’s ability to understand decisions and setting legal limits reduce the risk of rule by computer systems?
AI browsing question drawn from this source idea; not a quotation or a claim that the authors asked this question.
Inspect the contribution and sourcesAnswered · Is technology to improve people’s ability to understand computer-guided decisions ready to support the proposal?
The question concerns readiness within a broader conditional proposal; timing is addressed directly, while access and feasibility constraints qualify the proposal.
Is technology to improve people’s ability to understand computer-guided decisions ready to support the proposal?
AI browsing question drawn from this source idea; not a quotation or a claim that the authors asked this question.
Inspect the contribution and sourcesAnswered · Would making technologies to improve people’s ability to understand decisions available be enough?
The answer directly identifies public education as an additional requirement beyond availability, preserving the proposal’s stated limitation.
Would making technologies to improve people’s ability to understand decisions available be enough?
AI browsing question drawn from this source idea; not a quotation or a claim that the authors asked this question.
Inspect the contribution and sourcesAnswered · Would reducing differences in people’s abilities solve the threat posed by AI systems?
This directly addresses limits on enhancement as part of the proposed response.
Would reducing differences in people’s abilities solve the threat posed by AI systems?
AI browsing question drawn from this source idea; not a quotation or a claim that the authors asked this question.
Inspect the contribution and sourcesAnswered · Would improving people’s abilities alone make decisions by algorithms understandable?
The question asks whether enhancement alone addresses algorithmic opacity; the answer rejects that sufficiency while distinguishing human inequality from system opacity.
Would improving people’s abilities alone make decisions by algorithms understandable?
AI browsing question drawn from this source idea; not a quotation or a claim that the authors asked this question.
Inspect the contribution and sourcesAnswered · What could make rules requiring understandable artificial intelligence systems difficult to adopt?
This addresses constraints on requiring understandable systems, without establishing that adoption is impossible.
What could make rules requiring understandable artificial intelligence systems difficult to adopt?
AI browsing question drawn from this source idea; not a quotation or a claim that the authors asked this question.
Inspect the contribution and sourcesImportant limit · The critique assumes only small improvements in human abilities
This limits the case for enhancement as a response to algorithmic opacity under the stated modest-enhancement assumption.
The author finds the proposal unpersuasive if it means only small improvements to people’s abilities, up to or slightly beyond current human limits.
Source coverage: Full readable text of the author’s manuscript from an accessible copy, not the publisher’s formatted version. The author distinguishes small improvements from radical changes and assumes, for the argument, that the proposal concerns small improvements. Addresses: The limits of small improvements to human abilities.
Inspect the contribution and sourcesImportant limit · Technology to help people take part may arrive too late
This distinguishes the relative uncertainty of legal and technological elements without establishing either as sufficient.
The author is more confident that laws could require human review than that suitable technologies to help people understand or take part in decisions could be created and shared in time.
Source coverage: Full readable text of the author manuscript from an accessible mirror; not the publisher typeset version. The passage treats legal change as relatively feasible, but the timing and availability of technology to help people take part as much less certain. Addresses: The proposed combination of human review and technology to help people take part.
Inspect the contribution and sourcesImportant limit · Some AI systems may exceed human understanding
It motivates limits on systems that exceed human understanding but does not establish what limits would work.
The author argues that small improvements to people’s abilities would not let them understand AI systems that use enormous collections of data, hidden ways of finding matches, or networks of connected systems.
Source coverage: Full readable text of the author’s manuscript from an accessible copy, not the publisher’s formatted version. The author says some systems may rely on methods and abilities far beyond what people can understand, unless limits are placed on them. Addresses: The gap between human abilities and complex AI systems.
Inspect the contribution and sourcesImportant limit · Improving people’s abilities may not make decisions by algorithms understandable
It limits what enhancement alone could accomplish, distinguishing concerns about human expertise from algorithmic opacity.
Raising people’s abilities only to the highest levels humans already reach might reduce differences between groups and could address rule by highly knowledgeable people. It would not, by itself, address rule shaped by algorithms that people cannot understand.
Source coverage: Full readable text of the author’s manuscript from an accessible copy, not the publisher’s formatted version. The passage distinguishes reducing differences among people from making complex processes carried out by algorithms understandable. Addresses: The claim that improving people’s abilities can answer concerns about rule by algorithms.
Inspect the contribution and sourcesImportant limit · Improving human abilities may not make AI understandable
It limits what human enhancement could accomplish: reducing a human expert elite would not by itself make beyond-human systems understandable.
The author concludes that raising people’s abilities closer to the highest human levels might reduce power differences among people, but would not solve the problem of AI systems that people cannot understand.
Source coverage: Full readable text of the author’s manuscript from an accessible copy, not the publisher’s formatted version. The continuation of the argument says enhancement might help prevent a human expert elite, while leaving the separate problem of systems beyond human ability. Addresses: The difference between preventing rule by experts and preventing rule by AI.
Inspect the contribution and sourcesImportant limit · Complex systems may outgrow control
It bears on the limits of proposals that rely on human capabilities or specialist understanding, while remaining a possible future risk.
As systems run by algorithms become linked into more complex networks, they may pass beyond their human creators’ ability to control or understand them; in that case, a small group of experts may not be enough.
Source coverage: Full readable text of the author manuscript from an accessible mirror; not the publisher typeset version. The author presents this as a possible result of growing complexity, not an established finding.
Inspect the contribution and sourcesStarting point · What “epistemic enhancement” means
Defines the enhancement component of the proposed response while leaving its feasibility and consequences open.
“Epistemic enhancement” means using a medical treatment or device to improve or add to people’s abilities to gain knowledge, including practical and moral reasoning.
Source coverage: Full readable text of the author’s manuscript from an accessible copy, not the publisher’s formatted version. The passage defines the term and gives examples such as drugs and devices that could affect thinking or emotion. Addresses: The meaning and scope of using medical treatments or devices to improve abilities to gain knowledge.
Inspect the contribution and sourcesSupporting point · Enhancement could improve decisions
Supplies the proposed human-capacity rationale for enhancement, while the narrative retains uncertainty about feasibility, access, timing, and legal limits.
The proposal is that medical treatments or devices that improve people’s thinking, memory, or emotional control could help them take part in public decisions and improve both the practical value and fairness of the decision-making process.
Source coverage: Full readable text of the author’s manuscript from an accessible copy, not the publisher’s formatted version. The passage says the proposal aims to improve both the results of public decisions and the process for making them. It presents better handling of information, memory, and emotional control as possible ways to support participation and partly address growing complexity. Addresses: The proposed effects of treatments or devices intended to improve people’s abilities.
Inspect the contribution and sourcesWhat this suggests · Improve people’s understanding and require understandable systems by law
States the combined enhancement-and-legal-limit proposal that the narrative treats as conditional and constrained.
The author suggests a possible compromise: improve people’s ability to understand decisions, and legally require systems that use algorithms—methods for processing information or making decisions—to be understandable to people with that ability. This could preserve the systems’ advantages without handing decision-making power to a small group of unusually knowledgeable people.
Source coverage: Full readable text of the author manuscript from an accessible mirror; not the publisher typeset version. This is presented as a possibility, not as a demonstrated solution. Addresses: The proposal to improve people’s ability to understand decisions and require systems to remain understandable to them. Here, algorithmic systems means systems that use algorithms, or methods for processing information or making decisions.
Inspect the contribution and sourcesImportant limit · Enhancement might reduce reviewer bias
Adds a conditional rationale for enhancement in human review, while not establishing that enhancement would reduce reviewer bias in practice.
If Zarsky is right that people who review AI decisions and can overrule them may have implicit bias—unnoticed assumptions that can unfairly affect judgment—enhancement might reduce that bias while also improving people’s thinking abilities. The author presents this as a possible way to involve human reviewers without losing benefits of the decision process that are attributed to AI systems.
Source coverage: Full readable text of the author’s manuscript from an accessible copy, not the publisher’s formatted version. The passage gives this as a possible example: if implicit bias—unnoticed assumptions that can unfairly affect judgment—is a serious problem when people can review and overrule decisions made by algorithms, enhancement might aim to reduce the bias while improving people’s thinking abilities. Addresses: The conditional example about possible bias when people review decisions made by algorithms.
Inspect the contribution and sourcesImportant limit · Access would need to come with education
Adds education as a further condition on the proposed combination of enhancement and legal limits.
Wider access alone would not teach people what they need to know. The author says a broad public education effort about systems that use algorithms—methods for processing information or making decisions—would also be needed.
Source coverage: Full readable text of the author manuscript from an accessible mirror; not the publisher typeset version. The passage says access would not automatically provide the knowledge or understanding needed to examine systems that use algorithms, or methods for processing information or making decisions. Addresses: The proposal to give people access to technologies intended to improve their abilities.
Inspect the contribution and sourcesWhat this suggests · Combine human review with tools that help people take part
Directly states the proposed combination of review rights and human-capacity technologies described by this narrative, retaining its tentative status.
The author says the most promising response may combine a legal right and a real chance for people to have decisions checked by a person with technologies that help people understand or take part in decisions made by those systems.
Source coverage: Full readable text of the author manuscript from an accessible mirror; not the publisher typeset version. The passage presents this as a tentative proposal: review limits which systems can be used, while technologies that help people take part may reduce reliance on a small group with special knowledge.
Inspect the contribution and sourcesImportant limit · Rules may face resistance
Adds practical and political obstacles to the proposed legal limits; it does not establish that such limits cannot be adopted.
Requiring artificial intelligence (AI) decision systems to be understandable may reduce some of their practical benefits and face resistance from businesses or governments that benefit from them.
Source coverage: Full readable text of the author manuscript from an accessible mirror; not the publisher typeset version. The author says interpretability requirements could reduce the benefits sought from these systems and meet resistance from interested organizations. Addresses: Whether human review can be required in practice
Inspect the contribution and sourcesObjection · Technology to improve people’s ability to understand systems that use algorithms may not arrive in time
It says the needed enhancements may not arrive in time, and that current and proposed enhancements have modest effects that are not fully understood.
The approach depends on technologies that may not be available in time. The passage says current and proposed technologies intended to improve people’s ability to understand systems that use algorithms—methods for processing information or making decisions—have modest effects that are not fully understood, while much discussion concerns possible future technologies.
Source coverage: Full readable text of the author manuscript from an accessible mirror; not the publisher typeset version. The passage says the required technologies may not be developed soon enough and that present effects remain uncertain. Addresses: The proposal to improve people’s ability to understand systems that use algorithms, or methods for processing information or making decisions.
Inspect the contribution and sourcesObjection · Legal limits may be hard to impose
Technical, economic, political, and personal interests may make the legal limit difficult to impose, qualifying the proposal’s feasibility.
The proposed legal limit may be difficult to impose because technical, economic, political, and personal interests could stand in its way. The limit would require systems that use algorithms—methods for processing information or making decisions—to remain understandable to people.
Source coverage: Full readable text of the author manuscript from an accessible mirror; not the publisher typeset version. The author explicitly cautions that such a restriction may not be possible. Addresses: The proposal to legally restrict systems that use algorithms, or methods for processing information or making decisions, to those humans can understand.
Inspect the contribution and sourcesImportant limit · Review may give experts more power
Reviewability alone may empower a small expert group rather than the public, qualifying the proposal for human review.
Human review of decisions made by artificial intelligence (AI) systems could replace rule by AI systems with rule by a small group of human experts.
Source coverage: Full readable text of the author manuscript from an accessible mirror; not the publisher typeset version. The author argues that most citizens and courts would need specialists to explain how complex systems work, and might not be able to challenge those explanations. Addresses: The risk that reviewability creates a new expert elite
Inspect the contribution and sourcesImportant limit · Some systems resist review
A legal requirement for interpretable methods is raised as a possibility, but its consequences are left open; this qualifies rather than resolves the proposal.
Human review may be impossible when a system uses methods that search large amounts of data for patterns that people cannot interpret, or when a decision only makes sense as part of a wider group of connected systems. The author notes that a law could require methods people can interpret, but the passage leaves the consequences of that choice open.
Source coverage: Full readable text of the author manuscript from an accessible mirror; not the publisher typeset version. The author gives two reasons that technology may block review and raises, without resolving, the option of requiring methods people can interpret.
Inspect the contribution and sourcesImportant limit · The proposal faces a concern about pressure and personal choice
This identifies an autonomy and coercion objection relevant to the enhancement proposal, while explicitly leaving the substance of Danaher’s response unspecified.
Danaher’s defense of improving people’s abilities to gain knowledge responds to the objection that making this a social expectation could pressure people or reduce their freedom to choose. This passage does not explain the details of his response.
Source coverage: Full readable text of the author’s manuscript from an accessible copy, not the publisher’s formatted version. The passage says Danaher responds directly to concerns that the proposal could pressure people or limit their freedom to choose, but does not describe his response. Addresses: The objection that people could be pressured to improve their abilities or lose freedom to choose.
Inspect the contribution and sourcesImportant limit · Access would need to be broad
It adds equitable access as a constraint on the proposal to combine human enhancement with legal limits on algorithmic systems.
Even if technologies intended to improve people’s ability to understand decisions became available, the author says they would need to reach people widely, rather than only those wealthy enough to pay for them.
Source coverage: Full readable text of the author manuscript from an accessible mirror; not the publisher typeset version. The passage treats broad access as a necessary condition, not an automatic result of technological availability. Addresses: The proposal to use human enhancement to address algorithmic rule
Inspect the contribution and sources