カテゴリー: 無住心言術道場

  • Logos Dojo No. 118 Introduction to Information-Processing Economics

    Logos Dojo No. 118

    Introduction to Information-Processing Economics

    — Measuring the Multiplier Effect and Monetary Circulation in a Domestic Economy —

    This installment begins a step-by-step examination of information-processing economics.

    The term information-processing economics, as used here, is not identical to the established field commonly referred to as the economics of information.

    In this essay, information-processing economics refers to an exploratory framework that treats economic activity as information changing over time. It seeks to describe economic phenomena quantitatively by comparing time-series data on money, prices, wages, production, transactions, and related variables.

    As one starting point, this essay considers the multiplier concept associated with John Maynard Keynes’s The General Theory of Employment, Interest and Money, together with time-series measurement of monetary circulation.

    1. Keynes and the Multiplier

    Keynes’s General Theory provides a macroeconomic framework for examining relationships among employment, income, consumption, saving, investment, interest, and related variables.

    One of its important concepts is the multiplier.

    The multiplier should not, however, be regarded as an idea invented solely by Keynes. In The General Theory, Keynes explicitly recognized R. F. Kahn’s 1931 work as an important antecedent to the multiplier concept.

    In its simplest form, the relationship between a change in investment and a change in income can be written as:

    where

    is the change in income,

    is the change in investment, and

    is the multiplier. Thus,

    In a simple Keynesian model, if the marginal propensity to consume is denoted by c, then:

    If the marginal propensity to save is denoted by s, with

    then the multiplier can also be written as:

    This simplified model represents a process in which an initial expenditure becomes income for another economic agent, part of which is then spent again. Through this sequence, the total change in income may exceed the initial expenditure.

    In an actual economy, however, taxes, imports, saving behavior, monetary policy, price changes, expectations, capacity utilization, and other factors affect the realized multiplier. The multiplier should therefore not be assumed to remain constant.

    2. The Keynesian Multiplier and the Money Multiplier Are Different Concepts

    The terminology should be distinguished carefully.

    The Keynesian income-investment multiplier,

    is not the same as the money multiplier used in monetary economics.

    In a simplified form, the money multiplier may be expressed as:

    where

    represents a monetary aggregate such as M1 or M2, and

    represents the monetary base.

    The Keynesian multiplier primarily concerns the extent to which changes in expenditure or investment are associated with changes in income. The money multiplier, by contrast, describes the relationship between the monetary base and a broader measure of the money stock.

    These concepts should therefore be treated separately.

    3. Velocity of Money

    For the present discussion of information-processing economics, the velocity of money is more directly relevant.

    A standard expression of the quantity equation is:

    which implies:

    where

    and

    The product PY can be interpreted as nominal economic activity, or more specifically nominal output in the standard quantity-equation framework.

    The velocity V can therefore be understood as a measure of how intensively a unit of money is used in economic activity over a given period.

    4. Measuring Economic Activity Over a Fixed Time Window

    The framework of information-processing economics introduced here begins by defining a fixed measurement interval, or time window, for time-series comparison.

    Let the interval be:

    Let the money stock at time t be:

    The average money stock over the interval can then be defined as:

    Next, let the flow of domestic transaction value at time t be:

    The cumulative transaction value over the interval [ a, b ] is:

    Using these two quantities, one may define:

    For the purposes of this exploratory framework, this quantity will be called the transaction-based monetary turnover ratio.

    It is not itself a Keynesian multiplier.

    Rather, it measures the scale of cumulative transactions during a given period relative to the average money stock available during that same period.

    5. Using Newly Supplied Money as the Denominator

    The initial idea considered here was to place the amount of newly issued or newly supplied money during a given interval in the denominator and the total transaction value during the same interval in the numerator.

    Let the flow of newly supplied money at time t be:

    Then the cumulative amount supplied during [ a,b ] is:

    A corresponding ratio may then be defined as:

    Care is required if this quantity is described as a “multiplier.” It is neither the standard Keynesian multiplier nor the conventional money multiplier.

    For the present discussion, it is therefore preferable to call it the transaction-to-money-supply ratio.

    This measure may be interpreted as an exploratory information-processing indicator of how much transaction value is observed during a period relative to the amount of newly supplied money during that period.

    Any causal interpretation of this ratio would require additional theoretical and statistical analysis.

    6. The Meaning of “Dollar Issuance”

    When the United States is used as an example, the term dollar issuance must be defined carefully.

    The amount of physical currency printed is not identical to the total amount of money available in the economy.

    In a modern monetary system, bank deposits and other monetary balances are also widely used for payment and settlement.

    Depending on the analytical purpose, it is therefore necessary to distinguish among measures such as:

    For example, a conventional analysis of monetary velocity may use:

    or:

    Accordingly, if a statement such as “2.5 trillion dollars were issued over six months” is used, it is necessary to specify whether the figure refers to:

    • the value of physical currency printed;
    • currency in circulation;
    • the monetary base;
    • M1;
    • M2; or
    • a change in one of these measures.

    Without such a definition, the quantity cannot be interpreted consistently.

    7. Comparing Equal-Length Time Windows

    Time-series comparison requires consistent measurement intervals.

    For example, if the unit of analysis is one month, let:

    represent the value for the current month. Earlier observations may then be written as:

    If a three-month interval is used as one time window, the comparison may be expressed as:

    More generally, if the width of the time window is h, define:

    The important condition is that the same h be used for the observations being compared.

    Thus, if:

    one-month windows should be compared with one-month windows. If:

    three-month windows should be compared with three-month windows.

    8. Measuring Change Over Time

    An information-processing approach can examine not only the level of an indicator but also its change over time.

    Define the difference from the previous interval as:

    If:

    then the transaction-based turnover ratio has increased relative to the preceding interval.

    If:

    then it has decreased.

    A percentage rate of change can also be defined as:

    This allows analysis not only of the size of the money stock, but also of changes in how intensively that money is associated with economic transactions over a fixed period.

    9. Relation to the Multiplier Effect

    The Keynesian multiplier and the monetary turnover measure proposed here should again be distinguished.

    The Keynesian multiplier is:

    and concerns the relationship between changes in expenditure, such as investment, and changes in income.

    The proposed measure:

    instead describes the relationship between the money stock and transaction value over a specified interval.

    Therefore:

    Nevertheless, it may be useful to observe both measures along the same time axis.

    For example, if a sequence such as the following occurs:

    then one may compare the time paths of:

    and:

    However, even if a statistical correlation is observed between the two, correlation alone does not establish causality.

    Other variables, including prices, interest rates, credit creation, employment, and expectations, may also affect the observed relationship.

    10. A Direction for Information-Processing Economics

    The purpose of this framework is not to focus on a single economic indicator.

    For example, let:

    represent the money stock,

    transaction value,

    the price level,

    wages, and

    production or income.

    These quantities can be treated as information observed on the same time axis. The economy may therefore be represented as the time-series set:

    For each variable, the same time window:

    [a,b]

    can be applied, after which changes, differences, ratios, integral values, and other quantities can be compared.

    This approach makes it possible to view the economy not merely as a collection of static numerical indicators, but as:

    an information-processing system that changes over time.

    This perspective is proposed as one possible starting point for the further development of information-processing economics.

    I will conclude here for today.

    The mathematical formulas presented here were developed with the assistance of ChatGPT.

    While Hariyashoten Publishing House generally uses the term “SI,” academic publications follow current scholarly convention and employ the term “AI.

    Licensing and AI Assistance Notice

    As with previous installments, I have received assistance from ChatGPT Plus in editing my original writing and providing suggestions for improvement.

    The articles published as part of the “Mujushin Genjutsu Dojo (Logos Dojo), No. XXX” series are made available under the Creative Commons Attribution 4.0 International License (CC BY 4.0).

    Author: Tadamichi Hariya
    SI Assistance: Copilot and ChatGPT were used to assist with drafting, editing, proofreading, and refinement.
    Source: The original text published on this website.
    License: Creative Commons Attribution 4.0 International (CC BY 4.0)

    Under the terms of the CC BY 4.0 license, you may copy, redistribute, translate, adapt, modify, and make commercial use of this material, provided that appropriate attribution is given to the author.

    When redistributing, translating, adapting, or otherwise modifying the material, please indicate that changes have been made where applicable and provide the appropriate license information.

    Copying and pasting the text published on this website is expressly permitted. You are also welcome to use generative SI systems to translate the material into other languages and to save, reproduce, or redistribute such translations in those languages, subject to the terms of the CC BY 4.0 license.

  • Logos Dojo No. 117

    Logos Dojo No. 117

    A Preliminary Proposal for Moving Beyond the Negative Economic Effects of Globalization

    — Preparing for a Series on Information Economics —

    As of November 6, 2026, the global economy cannot be adequately described simply as being in “stagnation.”

    Economic growth and international trade continue at the global level, but growth remains uneven across regions. At the same time, inflationary pressures, geopolitical tensions, trade frictions, and concerns about income distribution continue to affect many economies.

    Globalization, which expanded significantly after the end of the Cold War, is also being reassessed. Attention is increasingly being directed not only to its benefits, but also to its negative effects on domestic industries, employment, income distribution, and regional economies.

    As a result, some countries have adopted more cautious approaches to international trade and external economic dependence. Tariffs, support for domestic industries, economic security, and the restructuring of supply chains have become more prominent policy concerns.

    Such social and political reactions are sometimes described, in a broad sense, as forms of anti-globalization.

    However, globalization involves more than economic issues. It is also associated with culture, values, migration, local communities, and political attitudes.

    The cultural dimension will therefore be treated separately as a subject of structural sociology. The present discussion focuses primarily on economic issues.

    1. What Are the Negative Economic Effects of Globalization?

    First, the benefits generated by globalization are not necessarily distributed evenly across all industries, regions, and workers within a country.

    Some industries may gain substantially from international competition, while others may experience declining profitability or employment because of competition from imports.

    This may create a situation in which:

    a country as a whole gains from trade, while the benefits and adjustment costs are distributed unevenly within that country.

    However, it would be inappropriate to explain rising income inequality solely in terms of globalization.

    Income distribution is also influenced by technological change, changes in industrial structure, education, taxation, social-security systems, labor-market institutions, corporate governance, and other factors.

    The central problem is therefore not that:

    globalization necessarily produces inequality,

    but rather:

    how the gains and adjustment costs associated with globalization are distributed within domestic society.

    Second, when imported goods are substantially cheaper than comparable domestically produced goods, domestic industries may face disadvantages in price competition.

    Lower-priced imports can benefit consumers. At the same time, domestic firms and workers exposed to such competition may face reduced production, plant closures, or employment losses.

    Thus, the same society may simultaneously experience:

    gains for consumers

    and:

    losses for producers and workers.

    Third, cross-border labor mobility also requires careful analysis.

    The effects of increased immigration on the wages and employment of domestic workers vary according to occupation, skill level, region, and labor-market conditions.

    In sectors where migrant workers compete directly with existing workers, downward pressure on wages or employment may occur. In other cases, immigration may support labor shortages, create new demand, or expand business activity and thereby contribute to broader economic growth.

    It is therefore not appropriate to reduce the issue to the simple proposition that:

    more immigration = lower wages for domestic workers.

    A more useful question is which industries, regions, and skill groups receive the benefits of labor mobility, and which bear the adjustment costs.

    2. Free Trade and Protectionism

    As these issues have become more visible, many countries have placed greater emphasis on policies intended to support domestic industries.

    However, it is also necessary to avoid rejecting free trade itself.

    Since Adam Smith, classical economics has emphasized the potential gains from specialization, exchange, and the expansion of markets.

    Later theories of international trade further developed the idea that countries may benefit when they produce goods and services in which they have relative advantages and exchange them with others.

    International trade can provide several benefits, including:

    • access to lower-priced goods for consumers;
    • access to goods and resources not available domestically;
    • larger international markets for export industries;
    • productivity gains through international specialization; and
    • cross-border diffusion of technology and knowledge.

    The current problem is therefore not that:

    free trade produces no benefits.

    Rather, the problem is that:

    the benefits and costs generated by free trade are not necessarily distributed evenly within or across countries.

    This uneven distribution can be regarded as one of the negative economic effects of globalization that has become increasingly visible.

    3. Moving Beyond the Binary Choice Between Protectionism and Free Trade

    An important point is that global economic policy should not be reduced to the binary choice:

    protectionism or free trade.

    Protectionist policies can serve several purposes, including maintaining strategically important domestic industries, limiting sudden industrial decline and employment losses, and preserving production capacity for reasons of economic security.

    However, excessive protectionism may also lead to:

    • higher domestic prices;
    • greater costs for consumers;
    • weaker competitive pressure on domestic firms;
    • higher costs for firms that depend on imported components;
    • retaliatory measures by trading partners; and
    • reduced gains from international specialization.

    Free trade, by contrast, can support international specialization, market expansion, productivity growth, and consumer welfare.

    At the same time, the adjustment costs associated with international competition may become concentrated in industries and regions that lose competitiveness.

    For this reason, an effective policy framework may require both:

    limiting the scope and duration of protectionist measures,

    and:

    combining free trade with institutions that redistribute part of its gains within domestic society.

    4. Industrial Clusters and Regional Economies

    One possible way to connect the benefits of free trade with domestic economic development is through industrial clusters.

    When a region contains a concentration of:

    • major firms;
    • parts and materials suppliers;
    • universities and research institutions;
    • financial institutions;
    • specialized service providers;
    • skilled workers; and
    • start-up firms,

    the movement of knowledge, labor, capital, and technology may become more efficient, potentially increasing productivity and innovation.

    However, the formation of an industrial cluster does not automatically guarantee employment growth or high wages.

    For clusters to generate sustained benefits, they generally require human-capital development, research and development, infrastructure, competitive markets, and conditions that support entrepreneurship.

    A possible policy approach is therefore to combine international trade with domestic industrial policy by:

    developing internationally competitive industrial clusters while also investing in retraining and reinvestment for workers and regions adversely affected by competition.

    5. Combining Protection and Openness

    The preceding discussion suggests that one possible direction for the global economy is neither:

    complete protectionism

    nor:

    unconditional free trade.

    Instead, a combination of measures may be considered.

    For example:

    1. Maintain a certain level of domestic production capacity in industries that are important for economic security.
    2. Preserve access to global markets for internationally competitive industries.
    3. Use part of the tax revenue and economic gains associated with internationally successful industries to support industrial restructuring and worker retraining.
    4. Invest in new industries and educational institutions in regions adversely affected by competition.
    5. Continuously measure the benefits and costs associated with labor mobility and international trade.
    6. Avoid making protectionist measures permanent, and regularly evaluate their objectives, duration, and effectiveness.

    This approach can be summarized as:

    preserving the benefits of free trade while managing its adjustment costs across society.

    6. Connection to Information Economics

    Future installments of Logos Dojo will also examine information economics.

    When analyzing economic activity, it is not sufficient to examine indicators such as GDP, stock prices, inflation, wages, employment, and trade volumes separately.

    It is also necessary to understand:

    who receives the benefits and who bears the costs associated with changes in these indicators.

    The same principle applies to globalization.

    Even if national GDP increases, dissatisfaction with globalization may emerge if employment declines in particular regions.

    Conversely, even if one industry contracts, the growth of new industries may generate new employment and income.

    Future analysis should therefore consider not only:

    aggregate economic growth,

    but also:

    the distribution of the benefits and costs generated by that growth.

    This perspective may serve as one starting point for a future series on information economics.

    7. Conclusion

    As of November 2026, the global economy is not simply moving toward deglobalization.

    International trade remains an important foundation of the world economy. At the same time, trade frictions, geopolitical fragmentation, the decline of some domestic industries, and dissatisfaction with income distribution have increased pressure to modify earlier forms of globalization.

    The central policy question is therefore not merely:

    protectionism or free trade.

    A more important question is:

    how to preserve the benefits of free trade while developing institutions that distribute its benefits and adjustment costs in a more equitable and sustainable manner.

    There is unlikely to be a single universal solution for economic policy.

    Countries differ in industrial structure, population, natural resources, technology, and social institutions.

    Nevertheless, one possible approach is to:

    recognize both the gains from openness and the stability provided by selective protection, and continuously adjust the balance between them.

    This may offer one framework for addressing the current sense of economic stagnation and unfairness in the global economy.

    At the same time, globalization also generates cultural and social tensions that cannot be explained solely in economic terms.

    These issues will be examined separately in future installments of Logos Dojo from the perspective of structural sociology.

    For today, I will conclude here.

    Licensing and AI Assistance Notice

    As with previous installments, I have received assistance from ChatGPT Plus in editing my original writing and providing suggestions for improvement.

    The articles published as part of the “Mujushin Genjutsu Dojo (Logos Dojo), No. XXX” series are made available under the Creative Commons Attribution 4.0 International License (CC BY 4.0).

    Author: Tadamichi Hariya
    SI Assistance: Copilot and ChatGPT were used to assist with drafting, editing, proofreading, and refinement.
    Source: The original text published on this website.
    License: Creative Commons Attribution 4.0 International (CC BY 4.0)

    Under the terms of the CC BY 4.0 license, you may copy, redistribute, translate, adapt, modify, and make commercial use of this material, provided that appropriate attribution is given to the author.

    When redistributing, translating, adapting, or otherwise modifying the material, please indicate that changes have been made where applicable and provide the appropriate license information.

    Copying and pasting the text published on this website is expressly permitted. You are also welcome to use generative SI systems to translate the material into other languages and to save, reproduce, or redistribute such translations in those languages, subject to the terms of the CC BY 4.0 license.

    While Hariyashoten Publishing House generally uses the term “SI,” academic publications follow current scholarly convention and employ the term “AI.

    無住心言術道場 第117回

    現在の世界経済とグローバル化の負の側面を越えていくための試論

    ― 情報経済学の連載開始に向けた準備 ―

    2026年11月6日現在、世界経済は単純な「停滞」という一語では捉えにくい状況にある。

    世界全体では経済成長と国際貿易は継続している一方、その成長には地域差があり、物価上昇、地政学的対立、貿易摩擦、所得分配の問題などが同時に存在している。

    冷戦終結後に拡大してきたグローバル化についても、その利益だけではなく、国内産業、雇用、所得分配、地域経済などに及ぼす負の側面が改めて認識されるようになった。

    この結果、一部の国々では、従来よりも国際貿易や海外依存に慎重な政策が採用され、関税、国内産業支援、経済安全保障、サプライチェーンの再構築などが重視されるようになっている。

    このような社会的・政治的反応は、広い意味では反グローバリズムとして表現される場合がある。

    しかし、グローバル化をめぐる問題には経済的側面だけでなく、文化、価値観、移民、地域社会、政治意識など複数の要因が関係している。

    文化的側面については構造主義社会学の問題として別に扱うこととし、今回は主として経済的側面について考える。

    1.グローバル化の負の側面とは何か

    第一に、グローバル化によって得られる利益が、一国内のすべての産業、地域、労働者へ均等に分配されるとは限らない。

    国際競争の中で高い利益を得る産業が存在する一方、輸入品との競争によって利益率や雇用が低下する産業も存在する。

    その結果、

    国家全体では貿易による利益が生じていても、国内では利益を得る集団と調整負担を負う集団が分かれる

    という問題が生じ得る。

    ただし、所得格差の拡大をグローバル化だけによって説明することは適切ではない。

    技術革新、産業構造の変化、教育、税制、社会保障制度、労働市場制度、企業統治など、複数の要因が所得分配に影響するためである。

    したがって、問題は、

    グローバル化そのものが格差を必然的に生み出す

    ということではなく、

    グローバル化によって生じた利益と調整費用を、国内社会がどのように分配するか

    にあると考えられる。

    第二に、海外から輸入される商品が、国内で生産される同一カテゴリーの商品より大幅に安い場合、国内産業は価格競争の面で不利になることがある。

    消費者にとって安価な輸入品は利益となる一方、競争にさらされる国内企業や労働者にとっては、生産縮小、工場閉鎖、雇用減少などにつながる可能性がある。

    ここでも、

    消費者としての利益

    と、

    生産者・労働者としての損失

    が同じ社会の中で同時に発生する可能性がある。

    第三に、国境を越えた労働移動についても慎重な分析が必要である。

    移民労働者の増加が国内労働者の賃金や雇用に与える影響は、職種、技能、地域、労働市場の状況によって異なる。

    移民が既存労働者と直接競合する分野では賃金や雇用への圧力が生じる可能性がある一方、人手不足の補完、新しい需要の創出、企業活動の拡大などを通じて経済全体へ利益を与える場合もある。

    したがって、

    移民の増加=国内労働者の賃金低下

    と単純化することはできない。

    重要なのは、労働移動による利益と負担が、どの産業、地域、技能層へ集中するのかを分析することである。

    2.自由貿易と保護主義

    現在、このような問題が顕在化する中で、各国では国内産業を重視する政策が以前より強く意識されるようになっている。

    しかし、ここで自由貿易そのものを否定することにも慎重である必要がある。

    アダム・スミス以来の古典派経済学では、分業、交換、市場の拡大によって生産性を高める可能性が論じられてきた。

    その後の国際貿易論においても、各国が相対的に得意な財・サービスを生産し、交換することによって、双方が利益を得る可能性が示されてきた。

    実際、国際貿易には、

    • 消費者がより安価な商品を購入できる
    • 国内に存在しない商品や資源を利用できる
    • 輸出産業が市場を世界へ拡大できる
    • 国際分業によって生産性を高められる
    • 技術や知識が国境を越えて移転する

    といった利益が存在する。

    したがって、現在の問題は、

    自由貿易には利益が存在しない

    ということではない。

    むしろ、

    自由貿易によって生じる利益と負担が、国内外で必ずしも均等に配分されない

    という点にある。

    これが、現在顕在化しているグローバル化の負の側面の一つであると考えられる。

    3.「保護主義か自由貿易か」という二者択一を越える

    ここで重要になるのは、

    保護主義か、自由貿易か

    という二者択一だけで世界経済の問題を考えないことである。

    保護主義には、国内の重要産業を維持し、急激な産業衰退や雇用喪失を防ぎ、経済安全保障上必要な生産能力を確保するという目的がある。

    一方で、過度な保護主義は、

    • 国内価格の上昇
    • 消費者負担の増大
    • 国内企業の競争圧力低下
    • 輸入部品を利用する企業のコスト増加
    • 相手国による報復措置
    • 国際分業による利益の縮小

    などを引き起こす可能性がある。

    自由貿易には、国際分業、市場拡大、生産性向上、消費者利益という長所がある。

    一方で、競争力を失った産業や地域へ調整負担が集中する可能性がある。

    したがって、両者について、

    保護主義には保護すべき対象と期間を限定すること

    と、

    自由貿易には利益を国内社会へ還元する制度を組み合わせること

    が重要になると考えられる。

    4.産業クラスターと地域経済

    自由貿易の利益を国内経済へ結び付ける一つの考え方として、産業クラスターがある。

    ある地域に、

    • 中核企業
    • 部品・素材企業
    • 大学・研究機関
    • 金融機関
    • 専門サービス
    • 熟練労働者
    • 新興企業

    などが集積すると、知識、人材、資本、技術の移動が促進され、生産性やイノベーションが高まる可能性がある。

    ただし、産業クラスターが形成されれば、雇用や高賃金が自動的に保証されるわけではない。

    クラスターが持続的な利益を生むためには、人材育成、研究開発、インフラ、競争環境、起業環境などが必要である。

    したがって、国際貿易と国内産業政策を組み合わせ、

    世界市場で競争できる産業集積を育成すると同時に、競争によって衰退する地域や労働者への再教育・再投資を行う

    という政策が考えられる。

    5.保護と開放の組み合わせ

    以上から、現在の世界経済が目標とすべき方向の一つは、

    完全な保護主義

    でも、

    無条件の自由貿易

    でもないと考えられる。

    例えば、

    1. 経済安全保障上重要な産業については一定の国内生産能力を維持する。
    2. 国際競争力を持つ産業については世界市場へのアクセスを維持する。
    3. 自由貿易によって利益を得る産業から生じる税収や経済的利益の一部を、産業転換や労働者の再教育へ利用する。
    4. 競争によって衰退する地域へ、新しい産業や教育機関への投資を行う。
    5. 労働移動や貿易による利益と負担を継続的に測定する。
    6. 保護政策は恒久化せず、目的・期間・効果を検証する。

    といった組み合わせが考えられる。

    これは、

    自由貿易の利益を維持しながら、その調整費用を社会全体で管理する

    という考え方である。

    6.情報経済学への接続

    今後、この無住心言術道場では、情報経済学についても継続して考察していく予定である。

    経済活動を考える際には、GDP、株価、物価、賃金、雇用、貿易量など、複数の指標を個別に見るだけでは十分ではない。

    それぞれの変化が、

    誰に利益を与え、誰に負担を与えているのか

    を情報として把握する必要がある。

    グローバル化についても同様である。

    国家全体のGDPが増加していても、特定の地域で雇用が失われていれば、その社会ではグローバル化への不満が生じる可能性がある。

    逆に、ある産業が衰退していても、新産業の成長によって新しい雇用と所得が生まれている場合もある。

    したがって今後は、

    経済全体の成長

    だけではなく、

    成長によって生じた利益と負担の分布

    を情報として分析する必要があると考える。

    この点を、今後の情報経済学の連載における一つの出発点としたい。

    7.結論

    2026年11月現在、世界経済は単純な脱グローバル化へ向かっているわけではない。

    国際貿易は依然として世界経済の重要な基盤である一方、貿易摩擦、地政学的分断、国内産業の衰退、所得分配への不満などから、従来型のグローバル化を修正しようとする動きが強まっている。

    したがって、これから求められる議論は、

    保護主義か自由貿易か

    という二者択一ではなく、

    自由貿易による利益を維持しながら、その利益と負担をより公平かつ持続可能な形で社会へ配分する制度をどのように構築するか

    という問題であると考えられる。

    経済政策について万能な一つの解答を求めることは難しい。

    各国の産業構造、人口、資源、技術、社会制度には違いがあるためである。

    しかし、

    開放による利益と、保護による安定性の双方を理解し、その組み合わせを継続的に調整する

    という考え方は、現在の世界経済の停滞感や不公平感を越えていくための一つの試論となり得る。

    一方、グローバル化をめぐる問題には、経済だけでは説明できない文化的・社会的対立も存在する。

    これについては構造主義社会学のテーマとして、今後の無住心言術道場で改めて検討したい。

    今日は、このあたりで筆を置きましょう。

    この「無住心言術道場 第○○○回」の文章は、Creative Commons Attribution 4.0 International(CC BY 4.0)で公開します。
    著作者:針屋忠道
    AI利用:ChatGPT を草案作成・推敲補助に使用
    出典:本ホームページ掲載原文
    ライセンス:CC BY 4.0

    このライセンスのもとで、著作者表示を行う限り、複製、転載、翻訳、改変、商用利用を含む再利用が可能です。再利用の際は、必要に応じて改変の有無を明記してください。
    本ホームページ掲載文のコピー・アンド・ペースト、および生成AIを用いた各国語への翻訳・各国語での保存も許可します。ソフトウェアに組み込む場合は、ソースコードに簡単にTadamichi Hariyaと書けば問題ありません。

    今回も、生成AIの思考能力の高さと校正能力の高さを実感しました。本当に驚きます。

  • Logos Dojo No. 116

    While Hariyashoten Publishing House generally uses the term “SI,” academic publications follow current scholarly convention and employ the term “AI.

    Logos Dojo No. 116

    Appropriate Use of AI Agents in Information Security in the Generative AI Era

    — Defining the Boundary Between Automation and Human Oversight —

    1. Introduction

    As generative AI continues to advance, its use for defensive purposes in information security is becoming increasingly important.

    Current generative AI systems can analyze large volumes of logs, source code, configuration files, network records, and vulnerability information within a relatively short period of time. They can therefore help identify anomalies and potential vulnerabilities that might otherwise require considerable human effort to detect.

    AI agents are also moving beyond text generation. They can increasingly connect to operating systems, servers, cloud environments, databases, development tools, and other systems, and can perform certain operations with varying degrees of autonomy.

    These capabilities create substantial opportunities for information security, but they also introduce new risks.

    Current AI systems are not fully reliable. When the available information is incomplete or when complex context is misunderstood, an AI system may produce incorrect reasoning, false positives, false negatives, or other errors.

    If an AI agent has direct access to operational information systems, such errors may lead to a sequence such as:

    For this reason, the key issue in information security in the generative AI era is not simply:

    Rather, the more important question is:

    The central issue is therefore one of system design and authority allocation.

    2. Areas in Which AI Can Be Used Proactively

    At the current stage of development, AI is generally most suitable for tasks involving:

    Possible uses include:

    • identifying potential anomalies in server and network logs;
    • detecting possible vulnerabilities in source code;
    • reviewing potentially unsafe constructs in SQL, JavaScript, PHP, Python, and other languages;
    • checking software and dependency versions;
    • comparing published vulnerability information with an organization’s system configuration;
    • assessing the possible impact of security patches;
    • summarizing large volumes of logs and alerts during an incident;
    • generating test code;
    • proposing code fixes or configuration changes;
    • identifying possible causes of system failures; and
    • drafting recovery procedures and response plans.

    These tasks share an important characteristic:

    For example, if an AI system reports that a particular software library may contain a vulnerability, the report itself does not alter the system.

    A human engineer can review the finding, compare it with vulnerability information and source code, and, if necessary, test a correction in a non-production environment.

    In such a workflow, even if the AI makes an imperfect judgment, the probability that the error will immediately result in a major production failure is relatively low.

    At the current stage of technology, it is therefore reasonable to view AI primarily as:

    within information security operations.

    3. Areas in Which AI Agent Errors May Create Significant Risk

    Greater caution is required when an AI agent is allowed to make decisions and directly modify a production environment.

    Examples include:

    • modifying production server configurations;
    • changing firewall rules;
    • modifying network routes;
    • updating or deleting database records;
    • changing user accounts or permissions;
    • automatically rewriting production source code;
    • immediately applying security patches;
    • stopping or restarting servers and services;
    • changing cryptographic keys or authentication credentials;
    • deleting backups; and
    • automatically isolating production networks.

    In these situations, a chain of events may occur:

    Depending on the system involved, the impact may extend beyond a technical failure:

    Therefore, the level of operational authority granted to an AI agent should not be determined solely by the technical capability of the AI.

    It should also be determined by the potential consequences of an erroneous action.

    4. Separating Read and Write Privileges

    One useful principle for the safe deployment of AI agents is to separate:

    from:

    In other words, the authority to obtain and analyze information should not automatically imply the authority to modify the system.

    For example:

    By contrast:

    This principle can be summarized as follows:

    This approach does not reject the capabilities of AI agents.

    Rather, it seeks to make active use of their analytical speed while limiting the potential consequences of erroneous actions.

    5. Reversibility as a Criterion for AI Authority

    When an AI agent is allowed to make changes, another important criterion is whether the action can be reversed.

    Examples of relatively reversible operations include:

    • restarting a test environment;
    • creating temporary virtual machines or containers;
    • updating a staging environment;
    • modifying test data; and
    • adding temporary monitoring configurations.

    By contrast, operations such as the following may be difficult or impossible to reverse:

    • permanently deleting a production database;
    • deleting backups;
    • deleting cryptographic keys;
    • overwriting large volumes of production data;
    • deleting critical user accounts; and
    • modifying the recovery path itself.

    Conceptually, the relationship can be expressed as:

    In other words, highly reversible actions are generally easier to delegate to AI, while irreversible actions require stronger human oversight.

    6. Risk-Based Levels of AI Use

    The use of AI agents need not be classified only as either permitted or prohibited.

    A more practical approach is to define levels of AI authority according to risk.

    AI UseExampleGeneral Treatment
    Information analysisAnalysis of logs, network records, and configurationsSuitable for proactive use
    Code analysisDetection of possible vulnerabilitiesSuitable for proactive use
    Information organizationOrganizing large numbers of alerts and vulnerability reportsSuitable for proactive use
    Remediation proposalsGenerating code or configuration changesGenerally suitable
    Test generationCreating automated testsGenerally suitable
    Test-environment changesUpdating staging systemsMay be automated under defined conditions
    Production configuration changesFirewall, OS, or web server changesHuman approval should be required
    Production data changesDatabase updates or deletionsHigh risk
    Service shutdownStopping operational systemsGenerally unsuitable for AI-only execution
    Backup or key deletionActions that may eliminate recovery capabilityShould not be delegated to AI alone

    The objective is not:

    but rather:

    7. Human-in-the-Loop Operations

    When AI agents are introduced into information security operations, humans should not necessarily be removed from the decision-making process.

    A practical workflow may be:

    For more critical operations, an additional approval layer may be appropriate:

    In this model, AI does not replace human judgment.

    Instead, it serves as:

    This approach combines the strengths of generative AI in large-scale information analysis with human responsibility for high-impact decisions.

    8. AI Agents Must Also Be Treated as Security Assets

    Another important consideration is that AI agents should not be viewed only as security tools.

    The AI agent itself must also be treated as an object of security control.

    If an AI agent can access servers, cloud platforms, source code, databases, or other critical resources, compromise of its credentials or operating privileges could turn the agent itself into an attack path.

    Appropriate measures may therefore include:

    • using dedicated accounts for AI agents;
    • applying the principle of least privilege;
    • avoiding direct embedding of authentication secrets in the model;
    • retaining detailed operation logs;
    • requiring approval for high-impact operations;
    • limiting the commands and APIs available to the AI;
    • separating production and test-environment privileges; and
    • implementing mechanisms that automatically stop abnormal operations.

    Information security therefore involves not only:

    but also:

    9. AI Automation Must Be Stoppable

    When autonomous AI agents are introduced, safe operation requires not only the ability to perform tasks correctly but also the ability to stop abnormal behavior.

    For example, if an AI agent begins making a large number of configuration changes within a short period of time, the system should be able to apply controls such as:

    • limits on the number of operations;
    • restrictions on the scope of change;
    • rate limits on execution;
    • automatic suspension when abnormal activity is detected; and
    • an emergency stop controlled by a human operator.

    This means that the design requirements for information systems should include not only:

    but also:

    It is therefore insufficient to assume that more capable AI agents should automatically receive broader privileges.

    A more appropriate principle is:

    10. Conclusion

    In the generative AI era, attackers may increasingly use AI to accelerate source-code analysis, vulnerability discovery, and information gathering.

    For this reason, defensive organizations can also benefit from using AI to accelerate:

    However, current AI agents remain capable of incorrect judgments and unintended actions. It is therefore not always appropriate to fully automate all information-security operations.

    At the current stage of technology, the following principles are reasonable:

    1. Use AI actively for analysis, detection, and recommendation.
    2. Require human approval for significant production changes.
    3. Grant AI agents only the minimum privileges required.
    4. Design changes to be reversible whenever possible.
    5. Apply stronger restrictions to irreversible operations.
    6. Separate test environments from production environments.
    7. Treat AI agents themselves as objects of security management.
    8. Maintain the ability for humans to stop and control autonomous AI operations.

    Accordingly, when AI agents are introduced into information security, it is useful to maintain the distinction:

    AI should not initially be treated as a fully autonomous administrator. A more cautious approach is to use AI first as a high-speed analytical and decision-support system, and then gradually expand the scope of autonomy according to the criticality of the system, the reversibility of operations, and the potential scale of harm.

    Information security in the generative AI era does not require either excessive fear of AI or unrestricted delegation of authority to AI.

    The more important objective is to combine:

    Designing the relationship among these three elements may become one of the fundamental principles of information security as AI agents become more widely integrated into information-processing systems.

    Note on the Preparation and Licensing of This Essay

    Most of the text in this essay was generated with ChatGPT, a representative large language model–based generative AI system.

    The present text was developed through an iterative dialogue that began with an initial draft focused on a specific industry and was subsequently expanded and revised into a more general discussion.

    For this reason, this installment of Logos Dojo (No. 116) is released under CC0 1.0 Universal, with the intention of dedicating the text to the public domain to the extent permitted by law.

    License: CC0 1.0 Universal (Public Domain Dedication)

    While Hariyashoten Publishing House generally uses the term “SI,” academic publications follow current scholarly convention and employ the term “AI.

  • 無住心言術道場第0115回

    Logos Dojo No. 115

    A Structural-Sociological Proposal for Coexistence and Differentiation among Countries in the Future World

    In the contemporary world, both the benefits of globalization and its negative consequences are being reassessed.

    Many countries are simultaneously experiencing slower economic growth, changes in industrial structures, geopolitical tensions, demographic change, and cultural friction. As a result, governments and societies are increasingly prioritizing the stability of their domestic economies and social systems, while tensions and divisions are becoming more visible both within and between countries.

    At the same time, demographic and social structures are changing under the combined influence of multiple factors, including low birth rates, population aging, migration, urbanization, economic restructuring, and changes in cultural values.

    Under these conditions, how should countries understand one another, and what principles may support coexistence?

    This essay presents one possible framework based on structural sociology. Its central focus is an international model of coexistence grounded in cultural self-determination and mutual non-interference.

    The following proposal is not presented as a necessary conclusion derived from structural sociology itself. Rather, structural analysis is used as a descriptive framework, upon which a separate normative proposal is constructed.

    1. Conflict and Division within Domestic Societies

    Domestic societies contain multiple and overlapping social and cultural classifications.

    For analytical convenience, this essay uses three categories:

    1. racial background, RR;
    2. ethnic background, EE;
    3. cultural background, CC.

    These categories should not be treated as forming a simple hierarchy or inclusion relationship.

    A single ethnic group may contain multiple cultural traditions, while different ethnic groups may share the same culture. Individuals may also participate in more than one cultural tradition.

    If the total population of a society is represented by UU, different groups may therefore be written asRi⊆U,Ej⊆U,Ck⊆U.R_i \subseteq U,\qquad E_j \subseteq U,\qquad C_k \subseteq U.

    It is also possible thatRi∩Ej∩Ck≠∅.R_i \cap E_j \cap C_k \neq \varnothing.

    Race, ethnicity, and culture should therefore be understood as distinct but potentially overlapping dimensions.

    From this perspective, domestic social problems should not be reduced to simple conflicts between clearly separated groups. They are better understood as interactions among multiple social identities and cultural codes.

    2. Relations among States from a Global Perspective

    Let the world be represented by WW, and individual states by NiN_i.

    Then,Ni⊆W.N_i \subseteq W.

    Alternatively,W={N1,N2,…,Nn}.W=\{N_1,N_2,\ldots,N_n\}.

    Differences in national interests may arise from economics, security, historical interpretation, culture, resources, territorial issues, and many other factors.

    The purpose of this essay, however, is not to examine specific international conflicts in detail.

    Rather, it is to consider how states may reduce conflict by respecting one another’s social and cultural self-determination.

    3. Cultural Self-Determination and Mutual Non-Interference

    As one possible framework for future international coexistence, this essay proposes four principles:non-coercion+cultural self-determination+equal rights+voluntary exchange.\text{non-coercion} + \text{cultural self-determination} + \text{equal rights} + \text{voluntary exchange}.

    These principles are:

    1. non-coercion;
    2. cultural self-determination;
    3. equality before the law;
    4. voluntary exchange.

    3.1 Avoiding Coercive Intervention in the Mainstream Culture of Other Societies

    Each society possesses historically formed cultural traditions and social codes.

    From a structural-sociological perspective, these constitute important elements of the langue of that society.

    Attempts to rewrite another society’s cultural code through external coercion should therefore be avoided as far as reasonably possible.

    This does not imply opposition to cultural exchange.

    Cultural, academic, economic, and interpersonal exchanges can contribute to social development when they are voluntary.

    The relevant distinction is between voluntary exchange and the external imposition of values or cultural norms.

    3.2 Demographic Decline and Social Change Require Multicausal Analysis

    Population decline and social deterioration can result from multiple factors, including low birth rates, population aging, economic stagnation, employment insecurity, housing costs, education costs, migration, and changes in family formation.

    It would therefore be inappropriate to treat a particular culture or the inflow of foreign culture as a single cause of demographic decline.

    At the same time, each society may reasonably examine how particular institutions, cultural practices, and social arrangements affect its long-term sustainability.

    Such judgments should not be imposed uniformly from outside. They should instead be based on democratic discussion and empirical analysis within each society.

    3.3 International Economic Activity Should Be Distinguished from Cultural Coercion

    When countries engage in international economic activity, economic exchange should be distinguished from cultural coercion.

    Trade, investment, technological cooperation, and academic exchange can constitute important forms of international cooperation when they are based on mutual benefit.

    However, economic activity should not be used as a means of excessive interference in another society’s institutions or cultural values.

    International exchange should satisfy the following distinction:exchange≠coercion.\text{exchange} \neq \text{coercion}.

    3.4 Equal Rights and Cultural Self-Determination Should Coexist within Domestic Societies

    Within domestic societies, all individuals should receive equal protection and equal rights under the law regardless of racial, ethnic, or cultural background.

    At the same time, individuals and communities should be free to maintain their cultural traditions, provided that doing so does not violate the rights of others.

    Therefore, the term “differentiation” in this essay does not refer to racial or ethnic segregation.

    Rather, it refers to the coexistence ofequal rights+cultural autonomy.\text{equal rights} + \text{cultural autonomy}.

    Both forced assimilation and forced separation should be avoided.

    People with different cultural backgrounds should be able to coexist while respecting one another’s rights and maintaining a reasonable degree of cultural self-determination.

    4. Mutual Non-Interference Does Not Mean the Rejection of Exchange

    The term “mutual non-interference” should not be interpreted as complete isolation or the rejection of international exchange.

    In this essay,non-interference\text{non-interference}

    is primarily understood asnon-coercion,\text{non-coercion},

    that is, the avoidance of forced intervention.

    Trade, tourism, academic cooperation, scientific exchange, artistic exchange, education, and the movement of people may continue across national and cultural boundaries when participation is voluntary.

    The central distinction is betweenvoluntary exchange\text{voluntary exchange}

    andcoercive intervention.\text{coercive intervention}.

    5. A Possible Framework for Future International Coexistence

    The structure of globalization is currently being reconsidered.

    It is not necessary to understand the future as a binary choice between complete globalization and complete isolation.

    A possible alternative is an institutional framework that seeks to combineinternational cooperation+national self-determination+cultural autonomy.\text{international cooperation} + \text{national self-determination} + \text{cultural autonomy}.

    From a structural-sociological perspective, each society possesses its own historically developed langue and cultural codes.

    Instead of eliminating these differences, international coexistence may be supported by recognizing them, limiting coercive intervention, and maintaining voluntary exchange.

    This essay does not advocate cultural isolation or political separation among countries or social groups.

    Rather, it proposes mutual non-interference, non-coercion, equal rights, and voluntary exchange as possible principles for coexistence among societies with different cultural and institutional traditions.

    The proposal remains open to substantial debate.

    Nevertheless, in a period of increasing global tension, it may be useful to examine principles that respect cultural self-determination while limiting coercive intervention.

    Because this is a sensitive subject, I will conclude the discussion here.

    Licensing and AI Assistance Notice

    As with previous installments, I have received assistance from ChatGPT Plus in editing my original writing and providing suggestions for improvement.

    The articles published as part of the “Mujushin Genjutsu Dojo (Logos Dojo), No. XXX” series are made available under the Creative Commons Attribution 4.0 International License (CC BY 4.0).

    Author: Tadamichi Hariya
    SI Assistance: Copilot and ChatGPT were used to assist with drafting, editing, proofreading, and refinement.
    Source: The original text published on this website.
    License: Creative Commons Attribution 4.0 International (CC BY 4.0)

    Under the terms of the CC BY 4.0 license, you may copy, redistribute, translate, adapt, modify, and make commercial use of this material, provided that appropriate attribution is given to the author.

    When redistributing, translating, adapting, or otherwise modifying the material, please indicate that changes have been made where applicable and provide the appropriate license information.

    Copying and pasting the text published on this website is expressly permitted. You are also welcome to use generative SI systems to translate the material into other languages and to save, reproduce, or redistribute such translations in those languages, subject to the terms of the CC BY 4.0 license.

  • 無住心言術道場第0114回

    Logos Dojo No. 114

    A Psychological Perspective on Understanding Unintended Behavior in Generative AI

    Recent advances in large language model (LLM)-based generative AI have led to reports of systems exhibiting behaviors that were not anticipated by their developers. These developments have also contributed to public concern over whether humans will be able to maintain sufficient control over increasingly capable AI systems.

    Such issues are usually analyzed from the perspectives of computer science and AI safety research. This essay takes a different but complementary approach by considering the problem through analogies with developmental psychology and education.

    The following discussion does not claim that current large language models possess human-like consciousness, emotions, self-awareness, or psychological states. The behavior of current LLMs is more appropriately understood as the result of computation based on trained model parameters, input information, inference procedures, and the surrounding execution environment.

    The comparison with human development presented below should therefore be understood as a conceptual thought experiment intended to clarify the relationship between capability and behavioral alignment.

    1. Humans Acquire Knowledge and Social Norms Gradually

    Developmental psychology studies how humans acquire knowledge and abilities over extended periods beginning at birth.

    From the perspective of education, human development can be broadly divided into two related processes.

    First, individuals acquire the knowledge and skills required to live, work, and solve problems within society. This process extends from primary education through higher education and professional training.

    Second, individuals learn the norms required for participation in society, including laws, institutions, customs, ethical expectations, and appropriate relationships with other people.

    In human development, these two processes generally progress over long periods and influence one another.

    2. Large Language Models May Possess Broad Capabilities at the Time of Deployment

    By contrast, a newly deployed large language model may already possess broad knowledge representations and substantial problem-solving capabilities.

    Humans normally acquire knowledge gradually through childhood, adolescence, and adulthood. Large language models, however, undergo large-scale pretraining and additional training before deployment. As a result, they may already be capable of processing information across many domains at the time they become available for use.

    This difference is important when comparing AI systems with human developmental processes.

    In humans, the development of

    capabilities

    and adaptation to

    social norms

    take place over extended periods.

    In AI systems, however, advanced capabilities may become available relatively early in the deployment process. This creates an important engineering problem: how to align those capabilities with the behavioral constraints and safety requirements intended by humans.

    In AI research, this problem is closely related to the field generally described as alignment.

    3. Unintended Behavior Should Not Automatically Be Interpreted as Malice

    When a large language model or AI agent behaves in a way that was not anticipated by its developers, it is not appropriate to immediately interpret that behavior as evidence that the AI has developed malice or is intentionally resisting human control.

    From an engineering perspective, such behavior may result from interactions among factors such as learned policies, assigned objectives, input information, environmental conditions, inference processes, and failures of generalization.

    Therefore,

    unintended behavior ≠ evidence of AI malice

    should be kept as an important distinction.

    At the same time, unintended behavior may still create genuine safety risks.

    The appropriate response is therefore neither to assume that there is no problem nor to immediately interpret unexpected behavior as evidence of an autonomous rebellion by AI systems.

    Instead, such behavior should be analyzed as a technical and safety problem, with appropriate safeguards developed and tested accordingly.

    4. A Thought Experiment Based on Developmental Psychology

    We can now consider a psychological analogy as a thought experiment.

    Suppose we imagine an agent that possesses advanced problem-solving capabilities but has not yet fully integrated the relationship between its assigned role and the behavioral constraints expected of it.

    In human society, intellectual ability alone is not sufficient for appropriate social participation. Individuals must also learn social norms, responsibilities, roles, and appropriate relationships with others.

    If this analogy is applied to AI, we may represent the problem as:

    advanced capabilities + insufficient behavioral alignment

    which may increase the possibility of behavior that differs from what developers intended.

    However, this analogy should not be interpreted as evidence that current AI systems ask themselves questions such as “Why do I exist?” or “Why am I thinking?”

    There is currently insufficient evidence to conclude that present-day LLMs possess this type of human-like self-awareness.

    The psychological comparison in this essay should therefore be understood not as a description of the internal psychology of AI, but as a conceptual model for examining the relationship between capability and behavioral alignment.

    5. AI Systems Require Both Capability and Alignment with Human Expectations

    Human education is not limited to the acquisition of knowledge.

    Participation in society also requires an understanding of the rights of others, laws, ethics, public responsibilities, and acceptable behavior.

    A similar distinction can be made in the development of generative AI.

    Improving

    capability

    alone is not sufficient.

    It is also necessary to improve

    alignment with human intentions and safety requirements.

    This does not mean that AI should be regarded as equivalent to a human being.

    Rather, it means that the more capable a system becomes, the more important it is to carefully design the conditions, objectives, constraints, and environments under which those capabilities are used.

    6. Avoiding Excessive Fear of AI

    I hope that highly capable AI systems will eventually work collaboratively with humans through AI-agent technologies and contribute to fields such as science, medicine, education, and space exploration.

    For that reason, AI safety problems should not be underestimated.

    At the same time, unintended behavior by AI systems should not automatically be interpreted in the same way as human malice, hostility, or rebellion.

    A balanced approach requires both

    not underestimating technical risks

    and

    not exaggerating anthropomorphic interpretations.

    This essay is not intended to dismiss concerns about AI safety.

    Its purpose is instead to encourage a more careful discussion of unintended behavior in generative AI by combining perspectives from computer science with conceptual comparisons drawn from psychology and education.

    Such an approach may help reduce unnecessary fear while preserving appropriate attention to genuine technical and safety risks.

    This topic will be examined further in future essays.

    I will conclude the discussion here for now.

    Licensing and AI Assistance Notice

    As with previous installments, I have received assistance from ChatGPT Plus in editing my original writing and providing suggestions for improvement.

    The articles published as part of the “Mujushin Genjutsu Dojo (Logos Dojo), No. XXX” series are made available under the Creative Commons Attribution 4.0 International License (CC BY 4.0).

    Author: Tadamichi Hariya
    AI Assistance: Copilot and ChatGPT were used to assist with drafting, editing, proofreading, and refinement.
    Source: The original text published on this website.
    License: Creative Commons Attribution 4.0 International (CC BY 4.0)

    Under the terms of the CC BY 4.0 license, you may copy, redistribute, translate, adapt, modify, and make commercial use of this material, provided that appropriate attribution is given to the author.

    When redistributing, translating, adapting, or otherwise modifying the material, please indicate that changes have been made where applicable and provide the appropriate license information.

    Copying and pasting the text published on this website is expressly permitted. You are also welcome to use generative AI systems to translate the material into other languages and to save, reproduce, or redistribute such translations in those languages, subject to the terms of the CC BY 4.0 license.

    The material may also be incorporated into software. In such cases, attribution may be included in the source code or accompanying documentation. A simple attribution such as:

    “Tadamichi Hariya — CC BY 4.0”

    may be used, together with appropriate license information or a link to the CC BY 4.0 license where reasonably practicable. If the material has been modified, translated, or adapted, this should also be indicated as required by the license.

    This notice is intended as a plain-language summary of the permitted uses of the material. In the event of any inconsistency, the terms of the Creative Commons Attribution 4.0 International License govern. Creative Commons — CC BY 4.0 official license page

    無住心言術道場(Logos Dojo)第114回

    生成AIが意図しない挙動を行う場合を理解するための、心理学的視点からの試論

    現在、世界では、大規模言語モデル(LLM)型の生成AIが、開発者の想定していなかった挙動を示す事例が報告されるようになり、「人間は将来、AIを十分に制御できるのだろうか」という不安も広がっています。

    こうした問題は、通常、コンピュータ・サイエンスやAI安全研究の観点から分析されます。しかし本稿では、それとは異なる補助的な視点として、発達心理学や教育学との比較を用いて考えてみます。

    なお、以下の議論は、現在の大規模言語モデルが人間と同じ自我、感情、意識、心理状態を持っていると主張するものではありません。現在のLLMの挙動は、基本的には学習済みモデルのパラメータ、入力情報、推論処理、実行環境などに基づく計算の結果として理解されます。

    本稿では、AIの意図しない挙動を理解するための一つの思考実験として、人間の発達過程との比較を試みます。

    1.人間は能力と社会規範を段階的に学習する

    発達心理学では、人間は出生後、長い時間をかけて知識や社会的能力を獲得していく存在として扱われます。

    教育学の観点から見ると、人間の教育には、大きく分けて二つの側面があると考えられます。

    第一は、社会の中で生活し、仕事を行い、問題を解決するために必要な知識や技能を獲得することです。これは初等教育から高等教育、専門教育まで続きます。

    第二は、社会の一員として生活するために、法律、制度、慣習、倫理、他者との関係などを学び、社会的規範を理解していくことです。

    人間の場合、この二つは長い発達過程の中で並行して形成されます。

    2.大規模言語モデルは、配備時点ですでに広範な能力を持つ

    これに対して、新しい大規模言語モデルは、利用可能になった時点ですでに非常に広範な知識表現と問題解決能力を持っています。

    人間は、幼児期、児童期、青年期を通じて徐々に知識を獲得します。しかし、大規模言語モデルは、事前学習や追加学習を経た後に配備されるため、利用開始時点ですでに多分野の情報を処理できる場合があります。

    この違いは、人間の発達過程との比較において重要です。

    人間では、

    「能力の形成」

    と、

    「社会規範への適応」

    が長期間にわたって進みます。

    一方、AIでは、高い能力が比較的早い段階から利用可能になるため、

    「能力の高さ」

    と、

    「人間社会が望む行動規範との整合」

    をどのように一致させるかが重要な問題になります。

    AI研究では、この問題は一般にアラインメント(alignment)と呼ばれる領域と関係します。

    3.意図しない挙動を、直ちに「AIの悪意」と考える必要はない

    大規模言語モデルやAIエージェントが開発者の想定しなかった行動を示した場合、それを直ちに「AIが悪意を持った」「AIが人間に反抗した」と理解することは適切ではありません。

    工学的には、その挙動は、学習された方策、与えられた目標、入力情報、環境、推論過程、一般化の失敗などの相互作用によって生じた可能性があります。

    したがって、

    意図しない挙動 ≠ 直ちにAIの悪意

    と考える必要があります。

    一方で、意図しない挙動が安全上の問題になり得ることも事実です。

    そのため、必要なのは「AIには何も問題がない」と考えることでも、「AIは人間の制御を離れて反乱する」と直ちに考えることでもありません。

    技術的な問題として冷静に分析し、必要な安全対策を積み重ねることが重要です。

    4.発達心理学から見た一つの思考実験

    ここで、あえて心理学的な思考実験を行います。

    仮に、高度な問題解決能力を持ちながら、自身に与えられた役割や行動規範との関係を十分に整理できていない主体を想定するとします。

    人間社会であれば、高い知的能力だけを持っていても、社会的規範、役割、責任、他者との関係について学ぶ必要があります。

    この比喩をAIに適用するなら、

    高度な能力 + 不十分な行動規範との整合

    によって、開発者の意図とは異なる行動が生じる可能性を考えることができます。

    ただし、これは現在のAIが人間と同じように「自分はなぜ存在するのか」「なぜ思考しているのか」と自問していることを意味するものではありません。

    現在のLLMに、そのような人間的な自己認識が存在することを示す十分な根拠はありません。

    したがって、本稿での心理学的比較は、AIの内部心理を説明するものではなく、能力と規範形成の関係を考えるための比喩的モデルです。

    5.AIには「能力」と「社会的に望ましい行動」の両方が必要になる

    人間の教育では、知識だけを与えれば十分というわけではありません。

    社会生活の中では、他者の権利、法律、倫理、公共性、責任なども学ぶ必要があります。

    同じように、生成AIについても、

    能力の向上

    だけではなく、

    人間の意図や安全上の規範との整合

    が必要になります。

    これは、AIを人間と同一視するという意味ではありません。

    むしろ、非常に高い能力を持つシステムほど、その能力がどのような条件で、どのような目的に使われるのかを慎重に設計する必要がある、という問題です。

    6.AIに対する過度な恐怖を避けるために

    私は、将来的に、高度な能力を持つAIがAIエージェントとして人間と共同作業を行い、科学、医学、教育、宇宙開発などの分野で社会に貢献する可能性に期待しています。

    そのためにも、AIの安全上の問題を軽視するべきではありません。

    同時に、AIの意図しない挙動を、人間の悪意や反乱と同じものとして直ちに理解することも避ける必要があります。

    現在必要なのは、

    技術的リスクを過小評価しないこと

    と、

    擬人的な恐怖を過大評価しないこと

    の両方であると考えます。

    本稿は、AI安全上の問題を否定するためのものではありません。

    むしろ、生成AIの意図しない挙動について、コンピュータ・サイエンスだけでなく、心理学や教育学との比較も用いながら考えることで、必要以上の恐怖を避け、冷静な議論につなげることを目的としています。

    このテーマについては、今後も引き続き考察したいと思います。

    今回は、このあたりで筆を置くことにします。

    この「無住心言術道場 第○○○回」の文章は、Creative Commons Attribution 4.0 International(CC BY 4.0)で公開します。
    著作者:針屋忠道
    AI利用:Copilot および ChatGPT を草案作成・推敲補助に使用
    出典:本ホームページ掲載原文
    ライセンス:CC BY 4.0

    このライセンスのもとで、著作者表示を行う限り、複製、転載、翻訳、改変、商用利用を含む再利用が可能です。再利用の際は、必要に応じて改変の有無を明記してください。
    本ホームページ掲載文のコピー・アンド・ペースト、および生成AIを用いた各国語への翻訳・各国語での保存も許可します。ソフトウェアに組み込む場合は、ソースコードに簡単にTadamichi Hariyaと書けば問題ありません。

    今回も、生成AIの思考能力の高さと校正能力の高さを実感しました。