Return-Path: Delivered-To: alvarochf@traceseis.com Received: from cs510.bluehost.com by cs510.bluehost.com with LMTP id IBF2MP+ysWhoiQYAWiBQxw (envelope-from ) for ; Fri, 29 Aug 2025 08:02:39 -0600 Return-path: Envelope-to: alvarochf@traceseis.com Delivery-date: Fri, 29 Aug 2025 08:02:39 -0600 Received: from m244-30.mailgun.net ([159.112.244.30]:63786) by cs510.bluehost.com with esmtps (TLS1.2) tls TLS_ECDHE_RSA_WITH_AES_128_GCM_SHA256 (Exim 4.98.2) (envelope-from ) id 1urzgd-00000001rHD-35R6 for alvarochf@traceseis.com; Fri, 29 Aug 2025 08:02:39 -0600 DKIM-Signature: a=rsa-sha256; v=1; c=relaxed/relaxed; d=mg-d1.substack.com; q=dns/txt; s=k1; t=1756476142; x=1756483342; h=List-Unsubscribe-Post: List-Unsubscribe: List-Post: List-Id: List-Archive: List-Owner: Reply-To: In-Reply-To: References: sender: sender: Date: Message-Id: To: To: From: From: Subject: Subject: Content-Type: Mime-Version; bh=2jaAfhV8FImda085xmXhzTzxiZUrk+vl42bEjtEtSC8=; b=iUaSori2nglxPiSdenlHBzeVDTphwF/ItPbzJot7WvwRedsmN/C3POXitfUlq5dCFDwZvD9vQBpIqWspz3QIG7x77QwvE3m3eCg8BCLg7T1rqFmdz00AHIFctzOCozb4UJtw/Yq70bWQ/Fbk2uOOy01q0xJMvtKrwUOhw8GV9Ek= X-Mailgun-Sid: WyJiYTg1MiIsImFsdmFyb2NoZkB0cmFjZXNlaXMuY29tIiwiMzliMDhjNiJd Received: by 405496fbeb77706e45452bc2403d48b7734668576ee0af3fe7315e022ace92ec with HTTP id 68b1b2ee9c647bca44cfa703; Fri, 29 Aug 2025 14:02:19 GMT X-Mailgun-Sending-Ip: 159.112.244.30 X-Mailgun-Batch-Id: 68b1b2ec2f156b4fb336b35d Mime-Version: 1.0 Content-Type: multipart/alternative; boundary="15b23085597008b40ca007e2a4455b4315ed3854435f5396e5586f5dba2c" Subject: Is algorithmic mediation always bad for autonomy? From: Cosmos Institute To: alvarochf@traceseis.com X-Mailgun-Tag: post X-Mailgun-Track-Clicks: false Message-Id: <20250829140138.3.dd117f0725431467@mg-d1.substack.com> Date: Fri, 29 Aug 2025 14:01:38 +0000 Feedback-ID: post-172159011:cat-post:pub-2225794:substack sender: Cosmos Institute References: In-Reply-To: Reply-To: Cosmos Institute List-Owner: List-URL: List-Archive: List-Id: List-Post: List-Unsubscribe: List-Unsubscribe-Post: List-Unsubscribe=One-Click X-Mailgun-Variables: {"category": "post", "email_generated_at": "1756476139320", "is_freemail": "true", "post_audience": "everyone", "post_id": "172159011", "post_type": "newsletter", "pub_community_enabled": "true", "publication_id": "2225794", "subdomain": "cosmosinstitute", "user_id": "347258985"} --15b23085597008b40ca007e2a4455b4315ed3854435f5396e5586f5dba2c Content-Type: text/plain; charset="utf-8" Content-Transfer-Encoding: quoted-printable View this post on the web at https://blog.cosmos-institute.org/p/is-algorit= hmic-mediation-always-bad Life is a navigation problem. There are too many places to go and too many = paths to take, too many things to try and too many skills to learn. For man= y of us, choice can be paralyzing. The ancients meditated on what sort of l= ife we ought to live, but today we wonder how best to spend our time when f= aced with an endless supply of ideas and experiences. One way we solve the = problem is by asking for help. We speak to our friends about the music they= like and browse lists on the internet about the best new books. We get recommendations that help us determine how best to spend our time. B= ut recommendations are no longer the preserve of humans; in fact, the vast = majority of them =E2=80=93 at least as it relates to media or travel =E2=80= =93 are already provided by machines. What movie to watch or what route to = take are questions answered by AI systems designed to keep you coming back = for more. These systems remind us that technology pervades our lives, that = we make decisions under conditions that we don=E2=80=99t always choose. On one level, this process is essential for the work of civilization. Alfre= d North Whitehead famously argued that society moves forward by =E2=80=9Cex= tending the number of important operations which we can perform without thi= nking about them.=E2=80=9D When I make coffee in the morning I don=E2=80=99= t think about grinding beans with a mortar and pestle; I just press a butto= n and wait for the machine to do its work. We can think about technology as the mechanism by which humans externalize = their capacities. In the 1967 book The Myth of the Machine, the American hi= storian Lewis Mumford argues [ https://substack.com/redirect/9b28176a-9217-= 406f-a756-236ccad297ea?j=3DeyJ1IjoiNXFxeXF4In0.h9dEDbGj64cTNudluPwbMYWklSu8= UKGoFv6eWS-HyoA ] technology can =E2=80=9Cselectively organize and consciou= sly direct both the internal and external agents of the mind.=E2=80=9D We m= ight say that technology is both instrumental (tools we use to achieve ends= ) and prosthetic (things we use to manifest our abilities in the world). Bu= t it also shapes the character of cognition, transforming our capacities as= they relate to the technological environment that we inhabit. Writing shifted the burden of recall from the rhythms of oral tradition to = marks in clay or ink. The abacus relieved the mind from the strain of calcu= lation, altering sequences of thought into patterns of beads that can be ph= ysically manipulated. Later, the compass and the map stretched human orient= ation across oceans to allow travellers to project themselves into spaces t= hey could not directly perceive. Technology extends human power into matter= so that action and thought can be carried further than the body or mind wo= uld allow, while unlocking new forms of cognitive activity that were previo= usly out of reach. For Plato, technology can be read as a form of craft knowledge, a disciplin= e that shaped both product and practitioner. The cobbler=E2=80=99s t=C3=A9c= hn=C4=93 produced shoes, but it also cultivated habits of judgment about fi= t, durability, and beauty; the navigator=E2=80=99s t=C3=A9chn=C4=93 guided = ships, but it also demanded an attunement to winds, stars, and currents. He= re technology is something like a training ground, a set of practices that = form the character of those who wield them. Technology externalizes human c= apacity, but it also bends those faculties back towards us by fostering new= dispositions and habits. Put differently: we make technology and technology makes us back. Making and breaking autonomy Autonomy can be understood as the cultivated capacity to deliberate well ab= out how to live, to revise one=E2=80=99s understanding through experience, = and to act on one's own judgment within a community that recognizes this sa= me capacity in others. Humans cultivate autonomy through practice and defen= d it against erosion, which requires the freedom to make choices and the di= scipline to reflect on their consequences. Technologies free attention for higher tasks while modifying the limits wit= hin which choices are made. Autonomy is bolstered when external aids streng= then the conditions for self-rule, but it begins to wither when those same = aids diminish the faculties they are meant to support. Every new tool is in= some sense both liberating and constraining, expanding our reach while alt= ering the kind of selves we are able to become. They benefit us when they m= aximise our capacity for deliberation, understanding, and judgement, and th= ey hamper us when they stymie these same faculties. We might say that technology mediates our interactions according to a gradi= ent of autonomy persevering (or autonomy degrading) forms: Technology that augments us. Technologies that augment us operate by extend= ing human faculties while leaving intact the basic work of judgment. They a= mplify what we can already do, stretching the reach of our bodies and minds= beyond their natural limits. Technology that configures us. Technologies that configure us arrange our f= ield of action. They shape the choices available, the order in which they a= ppear, and the ease with which one path can be taken over another. Technology that simulates us. Technologies that simulate us begin to imitat= e the processes of judgment we once reserved for ourselves. They generate c= andidate answers, options, or arguments that resemble the products of human= deliberation. Technology that replaces us. Technologies that replace us close the distanc= e between simulation and action. They no longer propose options for us to c= onsider but execute decisions on our behalf, carrying out tasks in a way th= at sidelines human deliberation. These categories describe dominant patterns, not exclusive kinds. All techn= ologies augment and configure us; the wheel augments locomotion while influ= encing settlement patterns and trade routes; the clock augments coordinatio= n while regimenting daily life. Some, like the abacus, may augment our math= ematical proficiency, configure economic exchange, and simulate working mem= ory. A human may still have to ratify the result, but in doing so these typ= es of technologies may dampen autonomy if the operator chooses to continual= ly outsource cognitive labor. Replacement is a different beast. Early flight instruments augmented the pi= lot=E2=80=99s senses and configured cockpit routines; later systems simulat= ed decision rules by offering candidate corrections for altitude or heading= =2E But once autopilot was installed, t= hose corrections were carried out auto= matically, with the pilot relegated to a supervisory role. Recommendation e= ngines may simulate taste, but autonomous content moderation agents go furt= her by making and executing decisions at speeds and scales that escape dire= ct oversight. But we know that pilots still know how to fly, and that human operators sti= ll set the rules for content-checking models and review decisions on reques= t. In this context, the faculty isn=E2=80=99t fully degraded so much as reo= rganized into new layers like writing rules, setting boundaries or reviewin= g edge cases (and we realise the benefit in terms of safer air travel or co= ntent-checking at scale). Autonomy contracts at the moment of execution but= reasserts itself in design and oversight. It persists in a new and differe= nt form. But is this enough for true autonomy? Returning to our definition, we could say that the problem with replacement= is that it interrupts the conditions under which autonomy is cultivated. T= o deliberate well, one has to practice deliberation; to revise one=E2=80=99= s understanding, one has to encounter feedback from experience; to act on o= ne=E2=80=99s judgment, one has to be allowed to make judgments in the first= place. Replacement is troubling because it risks producing agents who reta= in the status of autonomous individuals but lack the habits that autonomy r= equires. This is where technologies that replace autonomy depart from those that sim= ulate it. Unlike replacement, the act of simulation does not displace judgm= ent but reframes what it means to judge. When options arrive already filter= ed, scored, or composed, our role becomes one of selection rather than orig= ination. Simulation still lets us act on our own judgment =E2=80=94 the cho= ice is ours =E2=80=94 but it weakens the practice of deliberation. We don= =E2=80=99t wrestle with problems from first principles. Instead, we are han= ded plausible answers that shape the frame of thought before we begin. Feed= back from experience is still present, but it is feedback as it relates to = selection among simulations rather than on the process of generating judgme= nts ourselves. Systems thinking and thinking systems One way to think about whether or not a technology is autonomy persevering = is to compare it to other structures that mediate human activity. After all= , every one of us lives inside systems. Law channels disputes into courts, = education organises people through schools and curricula, and markets coord= inate activity through prices. Each system creates an environment where som= e decisions are easier, others harder, and some ruled out altogether. Max Weber=E2=80=99s account of bureaucracy articulates how [ https://substa= ck.com/redirect/a823a84f-d095-4f49-acae-82d7867be210?j=3DeyJ1IjoiNXFxeXF4In= 0.h9dEDbGj64cTNudluPwbMYWklSu8UKGoFv6eWS-HyoA ] systems of rules and proced= ures both enable and constrain human action. A bureaucracy strips away arbi= trariness by offering predictability and fairness, but it also narrows the = range of acceptable behavior in a manner that makes it harder to act outsid= e its prescribed channels. It draws into focus the double edge of systemati= zation: autonomy is enabled in the sense that individuals can plan and act = with security, but diminished in that those actions must follow the logic o= f the organization. If we apply this idea to our present moment, we might say that AI systems t= hat augment and configure judgement are better placed to preserve human aut= onomy than those that simulate or (in the most troubling instance) replace = it. That all said, configuration itself is a double-edged sword: in some ca= ses the limits on our choices are needed to avoid choice paralysis and prev= ent the most harmful outcomes, but equally some directed efforts at structu= ring from a third party (e.g., autoplay or default permissive privacy setti= ngs) risk infringing on autonomy. But which is AI? Artificial intelligence is a technology that takes the sha= pe of its container, a kind of thing that exists in a bewildering variety o= f contexts and configurations. The most commonly used is the recommender sy= stem, which underwrites the five hours a day that people spend on social me= dia [ https://substack.com/redirect/9a0444af-73fe-4eeb-bbed-ab21079ae38f?j= =3DeyJ1IjoiNXFxeXF4In0.h9dEDbGj64cTNudluPwbMYWklSu8UKGoFv6eWS-HyoA ], strea= ming video, [ https://substack.com/redirect/0448990e-9acf-4663-b111-42c2d2e= f5db9?j=3DeyJ1IjoiNXFxeXF4In0.h9dEDbGj64cTNudluPwbMYWklSu8UKGoFv6eWS-HyoA ]= and music platforms [ https://substack.com/redirect/087b2040-1e9a-4a5c-a40= 7-f70a3ebd3f5b?j=3DeyJ1IjoiNXFxeXF4In0.h9dEDbGj64cTNudluPwbMYWklSu8UKGoFv6e= WS-HyoA ] globally. Even assuming that only about two-thirds of this time i= s genuinely steered by recommendation engines, that tells us that roughly o= ne-fifth of our waking life is already mediated by algorithms. And this exc= ludes other domains where AI configures our behavior, from navigation syste= ms and online shopping to the automated curation of news feeds and adverts. Recommender systems simulate us by modelling our tastes and predicting what= we might want before we know it ourselves. They generate candidate options= =E2=80=94 be they songs, films or products =E2=80=94 that resemble the kin= ds of choices we would make had we the time or patience to think for oursel= ves. At the same time, these systems configure us by shaping the environmen= t in which decisions are taken. The catalogue of media is vast, so we are n= ecessarily provided with a handful of options that obey the system=E2=80=99= s logic of retention. Autonomy is shaped first by the simulations of what w= e might like, and second by the architectures that govern how and when thos= e simulations appear. Large language models are more ambiguous than recommender systems because t= heir primary effect seems to be to augment rather than constrain. They can = expand the reach of a researcher by drafting summaries, or stretch the rang= e of a programmer by generating code snippets. In these moments the technol= ogy amplifies human faculties, taking what we can already do and carrying i= t further without fixing the outcome in advance. Yet the same systems also = simulate us, producing judgments, arguments, and suggestions that imitate o= ur own reasoning. We might even say these outputs are representative of the= replacement model, depending on the extent to which users are scrutinizing= the final output. These types of systems can in principle be used to recommend anything you c= an think of. Songs, products, and films are all in play, but so too are tho= se related to how to think, how to respond, and how to live. The saving gra= ce is that users tend to directly ask language models for these recommendat= ions, rather than in the case of other types of recommender systems that te= nd to automatically provide suggestions without us consciously deliberating= about the kind of recommendation we want. But as models grow more capable, or become agents that can act directly in = the world, the systems become better placed to challenge human autonomy. An= AI model that not only drafts a contract but sends it, or not only writes = code but deploys it, has little need for human deliberation (except insofar= as it manifests in the oversight of these systems). Today, large models ar= e augmentative in theory but often simulative in practice. Tomorrow, they m= ay threaten the conditions under which autonomy is cultivated by making dec= isions with limited human input. Algorithmic futures Freedom is finite because we inhabit systems that guide our lives. That is = as true for us as it was for the ancients. Aristotle understood freedom as = the capacity to live well through the exercise of phron=C4=93sis, the pract= ical wisdom cultivated by experience. To deliberate wisely, one must encoun= ter situations, weigh alternatives, and learn from the consequences of one= =E2=80=99s choices. Autonomy, in this sense, grows as habits of reflection = are strengthened and withers when they are not. Technologies that augment o= r configure us can support this process by extending our capacities or free= ing us from thinking about the minutia of everyday life. But when technolog= ies begin to simulate or replace us, they risk undermining the very practic= es through which phron=C4=93sis is formed. Even Greek philosophers were mediated by technology: Aristotle=E2=80=99s Pe= ripatetic school was dependent on the scrolls, maps, and instruments that c= onfigured inquiry, giving shape to what could be known and transmitted. Phi= losophies of reason and the good life existed within the architectures of i= nscription and classification that made this mode of thought possible. Try = as we might, there is no escape from the systems that shape how we live in = the world. But AI represents a different challenge, with each move towards models that= do the thinking for us threatening to chip away at the foundations of auto= nomy. The path ahead involves recognizing its potential for structuring, si= mulating, and replacing our capacities, while reflecting on the types of th= ings we want to build. This is not a question of whether mediation itself i= s bad; we have, after all, always lived within structures that shape our ch= oices. Rather, the risk is that AI extends the powers of mediation into the= realm of deliberation itself. That=E2=80=99s the slow, difficult work of becoming the kind of person who = knows how they should live, a job that only one person is qualified for. Cosmos Institute [ https://substack.com/redirect/b96f48e3-9b7f-42b2-a7c9-03= 20d027a35e?j=3DeyJ1IjoiNXFxeXF4In0.h9dEDbGj64cTNudluPwbMYWklSu8UKGoFv6eWS-H= yoA ] is the Academy for Philosopher-Builders, with programs, grants, event= s, and fellowships for those building AI for human flourishing. Unsubscribe https://substack.com/redirect/2/eyJlIjoiaHR0cHM6Ly9ibG9nLmNvc21= vcy1pbnN0aXR1dGUub3JnL2FjdGlvbi9kaXNhYmxlX2VtYWlsP3Rva2VuPWV5SjFjMlZ5WDJsa0= lqb3pORGN5TlRnNU9EVXNJbkJ2YzNSZmFXUWlPakUzTWpFMU9UQXhNU3dpYVdGMElqb3hOelUyT= kRjMk1UTTRMQ0psZUhBaU9qRTNPRGd3TVRJeE16Z3NJbWx6Y3lJNkluQjFZaTB5TWpJMU56azBJ= aXdpYzNWaUlqb2laR2x6WVdKc1pWOWxiV0ZwYkNKOS5SQzlYWWM2X0JuSW5SSnVGUzhUdnBubFN= OQU1vWGhyNUJxb2k2RFFpN0VRIiwicCI6MTcyMTU5MDExLCJzIjoyMjI1Nzk0LCJmIjp0cnVlLC= J1IjozNDcyNTg5ODUsImlhdCI6MTc1NjQ3NjEzOCwiZXhwIjoyMDcyMDUyMTM4LCJpc3MiOiJwd= WItMCIsInN1YiI6ImxpbmstcmVkaXJlY3QifQ.tYyOgNNNwUwzwdmtAU0YL08w0gzfBCtXoakzJ= 4_LuZ4? --15b23085597008b40ca007e2a4455b4315ed3854435f5396e5586f5dba2c Content-Type: text/html; charset="utf-8" Content-Transfer-Encoding: quoted-printable Is algorithmic mediation= always bad for autonomy?3D""
We make technology, and technology makes us back
͏ =     ­͏     ­͏     = 3;͏     ­͏     ­͏   = 199; ­͏     ­͏     ­͏ &= nbsp;   ­͏     ­͏     ­= ;͏     ­͏     ­͏   Q= 99; ­͏     ­͏     ­͏ &n= bsp;   ­͏     ­͏     ­= ͏     ­͏     ­͏   ̳= 9; ­͏     ­͏     ­͏ &nb= sp;   ­͏     ­͏     ­&= #847;     ­͏     ­͏    = ; ­͏     ­͏     ­͏ &nbs= p;   ­͏     ­͏     ­&#= 847;     ­͏     ­͏    = ­͏     ­͏     ­͏  = ;   ­͏     ­͏     ­= 47;     ­͏     ­͏     = ­͏     ­͏     ­͏  =   ­͏     ­͏     ­T= 7;     ­͏     ­͏     &= #173;͏     ­͏     ­͏   =   ­͏     ­͏     ­͏= ;     ­͏     ­͏     &#= 173;͏     ­͏     ­͏   &= #8199; ­͏     ­͏     ­͏=     ­͏     ­͏     = 73;͏     ­͏     ­͏   &#= 8199; ­͏     ­͏     ­͏ =     ­͏     ­͏     = 3;͏     ­͏     ­͏   = 199; ­͏     ­͏     ­͏ &= nbsp;   ­͏     ­͏     ­= ;͏     ­͏     ­͏   Q= 99; ­͏     ­͏     ­͏ &n= bsp;   ­͏     ­͏     ­= ͏     ­͏     ­͏   ̳= 9; ­͏     ­͏     ­͏ &nb= sp;   ­͏     ­͏     ­&= #847;     ­͏     ­͏    = ; ­͏     ­͏     ­͏ &nbs= p;   ­͏     ­͏     ­&#= 847;     ­͏     ­͏    = ­͏     ­͏     ­͏  = ;   ­͏     ­͏     ­= 47;     ­͏     ­͏     = ­͏     ­͏     ­͏  =   ­͏     ­͏     ­T= 7;     ­͏     ­͏     &= #173;͏     ­͏     ­͏   =   ­͏     ­͏     ­͏= ;     ­͏     ­͏     &#= 173;͏     ­͏     ­͏   &= #8199; ­͏     ­͏     ­͏=     ­͏     ­͏     = 73;͏     ­͏     ­͏   &#= 8199; ­͏     ­͏     ­͏ =     ­͏     ­͏     = 3;͏     ­͏     ­͏   = 199; ­͏     ­͏     ­͏ &= nbsp;   ­͏     ­͏     ­= ;͏     ­͏     ­͏   Q= 99; ­͏     ­͏     ­͏ &n= bsp;   ­͏     ­͏     ­= ͏     ­͏     ­͏   ̳= 9; ­͏     ­͏     ­͏ &nb= sp;   ­͏     ­͏     ­&= #847;     ­͏     ­͏    = ; ­͏     ­͏     ­͏ &nbs= p;   ­͏     ­͏     ­&#= 847;     ­͏     ­͏    = ­͏     ­͏     ­͏  = ;   ­͏     ­͏     ­= 47;     ­͏     ­͏     = ­͏     ­͏     ­͏  =   ­͏     ­
<= tbody>
Forwarded this email? = Subscribe here for more

Is algorithmic mediation always bad for autonomy?

We= make technology, and technology makes us back

<= td style=3D"vertical-align:middle;">
Harry Law
=
 
=
3D""
3D""
3D""
READ IN A= PP3D""
 
Mikalojus Čiurlionis, Sonata of the Stars (1908)

Life is a na= vigation problem. There are too many places to go and too many paths to tak= e, too many things to try and too many skills to learn. For many of us, cho= ice can be paralyzing. The ancients meditated on what sort of life we ought= to live, but today we wonder how best to spend our time when faced with an= endless supply of ideas and experiences. One way we solve the problem is b= y asking for help. We speak to our friends about the music they like and br= owse lists on the internet about the best new books.

= We get recommendations that help us determine how bes= t to spend our time. But recommendations are no longer the preserve of huma= ns; in fact, the vast majority of them – at least as it relates to me= dia or travel – are already provided by machines. What movie to watch= or what route to take are questions answered by AI systems designed to kee= p you coming back for more. These systems remind us that technology pervade= s our lives, that we make decisions under conditions that we don’t al= ways choose.

On one level, this process is essential= for the work of civilization. Alfred North Whitehead famously argued that = society moves forward by “extending the number of important operation= s which we can perform without thinking about them.” When I make coff= ee in the morning I don’t think about grinding beans with a mortar an= d pestle; I just press a button and wait for the machine to do its work.

We can think about technology as the mechanism by which= humans externalize their capacities. In the 1967 book The Myth = of the Machine, the American historian Lewis Mumford argues technology can “selectively organize and consciously direct both t= he internal and external agents of the mind.” We might say that techn= ology is both instrumental (tools we use to achieve ends) and prosthetic (t= hings we use to manifest our abilities in the world). But it also shapes th= e character of cognition, transforming our capacities as they relate to the= technological environment that we inhabit.

Writing = shifted the burden of recall from the rhythms of oral tradition to marks in= clay or ink. The abacus relieved the mind from the strain of calculation, = altering sequences of thought into patterns of beads that can be physically= manipulated. Later, the compass and the map stretched human orientation ac= ross oceans to allow travellers to project themselves into spaces they coul= d not directly perceive. Technology extends human power into matter so that= action and thought can be carried further than the body or mind would allo= w, while unlocking new forms of cognitive activity that were previously out= of reach.

For Plato, technology can be read as a for= m of craft knowledge, a discipline that shaped both product and practitione= r. The cobbler’s téchnē produced shoe= s, but it also cultivated habits of judgment about fit, durability, and bea= uty; the navigator’s téchnē guided sh= ips, but it also demanded an attunement to winds, stars, and currents. Here= technology is something like a training ground, a set of practices that fo= rm the character of those who wield them. Technology externalizes human cap= acity, but it also bends those faculties back towards us by fostering new d= ispositions and habits.

Put differently: we make tec= hnology and technology makes us back.

Making and = breaking autonomy

Autonomy can be understood as t= he cultivated capacity to deliberate well about how to live, to revise one&= #8217;s understanding through experience, and to act on one's own judgment = within a community that recognizes this same capacity in others. Humans cul= tivate autonomy through practice and defend it against erosion, which requi= res the freedom to make choices and the discipline to reflect on their cons= equences.

Technologies free attention for higher tasks whil= e modifying the limits within which choices are made. Autonomy is bolstered= when external aids strengthen the conditions for self-rule, but it begins = to wither when those same aids diminish the faculties they are meant to sup= port. Every new tool is in some sense both liberating and constraining, exp= anding our reach while altering the kind of selves we are able to become. T= hey benefit us when they maximise our capacity for deliberation, understand= ing, and judgement, and they hamper us when they stymie these same facultie= s.

We might say that technology mediates our interactions a= ccording to a gradient of autonomy persevering (or autonomy degrading) form= s:

  • Technology that augments<= /strong> us. Technologies that augment us opera= te by extending human faculties while leaving intact the basic work of judg= ment. They amplify what we can already do, stretching the reach of our bodi= es and minds beyond their natural limits.

  • Technology that configures us. Technologies that con= figure us arrange our field of action. They shape the choices available, th= e order in which they appear, and the ease with which one path can be taken= over another.

  • Technology that simulates us. Technologies that simulate us begin to imitate th= e processes of judgment we once reserved for ourselves. They generate candi= date answers, options, or arguments that resemble the products of human del= iberation.

  • Technology that replaces us. Technologies that replace us close the distance betwee= n simulation and action. They no longer propose options for us to consider = but execute decisions on our behalf, carrying out tasks in a way that sidel= ines human deliberation.

These categories = describe dominant patterns, not exclusive kinds. All technologies augment a= nd configure us; the wheel augments locomotion while influencing settlement= patterns and trade routes; the clock augments coordination while regimenti= ng daily life. Some, like the abacus, may augment our mathematical proficie= ncy, configure economic exchange, and simulate working memory. A human may = still have to ratify the result, but in doing so these types of technologie= s may dampen autonomy if the operator chooses to continually outsource cogn= itive labor.

Replacement is a different beast. Early flight= instruments augmented the pilot’s senses and configured cockpit rout= ines; later systems simulated decision rules by offering candidate correcti= ons for altitude or heading. But once autopilot was installed, those correc= tions were carried out automatically, with the pilot relegated to a supervi= sory role. Recommendation engines may simulate taste, but autonomous conten= t moderation agents go further by making and executing decisions at speeds = and scales that escape direct oversight.

But we know that p= ilots still know how to fly, and that human operators still set the rules f= or content-checking models and review decisions on request. In this context= , the faculty isn’t fully degraded so much as reorganized into new la= yers like writing rules, setting boundaries or reviewing edge cases (and we= realise the benefit in terms of safer air travel or content-checking at sc= ale). Autonomy contracts at the moment of execution but reasserts itself in= design and oversight. It persists in a new and different form.

But is this enough for true autonomy?

R= eturning to our definition, we could say that the problem with replacement = is that it interrupts the conditions under which autonomy is cultivated. To= deliberate well, one has to practice deliberation; to revise one’s u= nderstanding, one has to encounter feedback from experience; to act on one&= #8217;s judgment, one has to be allowed to make judgments in the first plac= e. Replacement is troubling because it risks producing agents who retain th= e status of autonomous individuals but lack the habits that autonomy requires.

This is where technologies that replace autonomy depart from those that s= imulate it. Unlike replacement, the act of simulation does not displace jud= gment but reframes what it means to judge. When options arrive already filt= ered, scored, or composed, our role becomes one of selection rather than or= igination. Simulation still lets us act on our own judgment — the cho= ice is ours — but it weakens the practice of deliberation. We donR= 17;t wrestle with problems from first principles. Instead, we are handed pl= ausible answers that shape the frame of thought before we begin. Feedback f= rom experience is still present, but it is feedback as it relates to selection among simulations rather than on the process of g= enerating judgments ourselves.

Systems thi= nking and thinking systems

One way to think about= whether or not a technology is autonomy persevering is to compare it to ot= her structures that mediate human activity. After all, every one of us live= s inside systems. Law channels disputes into courts, education organises pe= ople through schools and curricula, and markets coordinate activity through= prices. Each system creates an environment where some decisions are easier= , others harder, and some ruled out altogether.

Max W= eber’s account of bureaucracy articulates how systems of rules= and procedures both enable and constrain human action. A bureaucracy strip= s away arbitrariness by offering predictability and fairness, but it also n= arrows the range of acceptable behavior in a manner that makes it harder to= act outside its prescribed channels. It draws into focus the double edge o= f systematization: autonomy is enabled in the sense that individuals can pl= an and act with security, but diminished in that those actions must follow = the logic of the organization.

If we apply this idea= to our present moment, we might say that AI systems that augment and confi= gure judgement are better placed to preserve human autonomy than those that= simulate or (in the most troubling instance) replace it. That all said, co= nfiguration itself is a double-edged sword: in some cases the limits on our= choices are needed to avoid choice paralysis and prevent the most harmful = outcomes, but equally some directed efforts at structuring from a third par= ty (e.g., autoplay or default permissive privacy settings) risk infringing = on autonomy.

But which is AI? Artificial intelligence= is a technology that takes the shape of its container, a kind of thing tha= t exists in a bewildering variety of contexts and configurations. The most = commonly used is the recommender system, which underwrites the five hours a= day that people spend on social media, streaming video, and = mus= ic platforms globally. Even assuming that only about two-thirds o= f this time is genuinely steered by recommendation engines, that tells us t= hat roughly one-fifth of our waking life is already mediated by algorithms.= And this excludes other domains where AI configures our behavior, from nav= igation systems and online shopping to the automated curation of news feeds= and adverts.

Recommender systems simulate us by mod= elling our tastes and predicting what we might want before we know it ourse= lves. They generate candidate options — be they songs, films or produ= cts — that resemble the kinds of choices we would make had we the tim= e or patience to think for ourselves. At the same time, these systems confi= gure us by shaping the environment in which decisions are taken. The catalo= gue of media is vast, so we are necessarily provided with a handful of opti= ons that obey the system’s logic of retention. Autonomy is shaped fir= st by the simulations of what we might like, and second by the architecture= s that govern how and when those simulations appear.

Large = language models are more ambiguous than recommender systems because their p= rimary effect seems to be to augment rather than constrain. They can expand= the reach of a researcher by drafting summaries, or stretch the range of a= programmer by generating code snippets. In these moments the technology am= plifies human faculties, taking what we can already do and carrying it furt= her without fixing the outcome in advance. Yet the same systems also simula= te us, producing judgments, arguments, and suggestions that imitate our own= reasoning. We might even say these outputs are representative of the repla= cement model, depending on the extent to which users are scrutinizing the f= inal output.

These types of systems can in principle be use= d to recommend anything you can think of. Songs, products, and films are al= l in play, but so too are those related to how to think, how to respond, an= d how to live. The saving grace is that users tend to directly ask language= models for these recommendations, rather than in the case of other types o= f recommender systems that tend to automatically provide suggestions withou= t us consciously deliberating about the kind of recommendation we want.

=

But as models grow more capable, or become agents that can act= directly in the world, the systems become better placed to challenge human= autonomy. An AI model that not only drafts a contract but sends it, or not= only writes code but deploys it, has little need for human deliberation (e= xcept insofar as it manifests in the oversight of these systems). Today, la= rge models are augmentative in theory but often simulative in practice. Tom= orrow, they may threaten the conditions under which autonomy is cultivated = by making decisions with limited human input.

Alg= orithmic futures

Freedom is finite because = we inhabit systems that guide our lives. That is as true for us as it was f= or the ancients. Aristotle understood freedom as the capacity to live well = through the exercise of phronēsis, the practical= wisdom cultivated by experience. To deliberate wisely, one must encounter = situations, weigh alternatives, and learn from the consequences of one̵= 7;s choices. Autonomy, in this sense, grows as habits of reflection are str= engthened and withers when they are not. Technologies that augment or confi= gure us can support this process by extending our capacities or freeing us = from thinking about the minutia of everyday life. But when technologies beg= in to simulate or replace us, they risk undermining the very practices thro= ugh which phronēsis is formed.

Even Greek philosophers were mediated by technology: Aristotle’= s Peripatetic school was dependent on the scrolls, maps, and instruments th= at configured inquiry, giving shape to what could be known and transmitted.= Philosophies of reason and the good life existed within the architectures = of inscription and classification that made this mode of thought possible. = Try as we might, there is no escape from the systems that shape how we live= in the world.

But AI represents a different challenge, wit= h each move towards models that do the thinking for us threatening to chip = away at the foundations of autonomy. The path ahead involves recognizing it= s potential for structuring, simulating, and replacing our capacities, whil= e reflecting on the types of things we want to build. This is not a questio= n of whether mediation itself is bad; we have, after all, always lived with= in structures that shape our choices. Rather, the risk is that AI extends t= he powers of mediation into the realm of deliberation itself.

That’s the slow, difficult work of becoming the kind of person w= ho knows how they should live, a job that only one person is qualified for.=

Cosmos Institute is the Academy for Philosoph= er-Builders, with programs, grants, events, and fellowships for those build= ing AI for human flourishing.

 <= td>
3D"=Like<= /span>
<= tr><= /tr>
3D""Comment
3D=Rest= ack
 

© 2025 = Cosmos Institute
500 W 2nd Street, Suite 1900, Austin, TX 7= 8701
Uns= ubscribe

3D"Get

<= /div>
3D""=3D"" --15b23085597008b40ca007e2a4455b4315ed3854435f5396e5586f5dba2c--