<?xml version="1.0" encoding="UTF-8"?><rss xmlns:dc="http://purl.org/dc/elements/1.1/" xmlns:content="http://purl.org/rss/1.0/modules/content/" xmlns:atom="http://www.w3.org/2005/Atom" version="2.0" xmlns:cc="http://cyber.law.harvard.edu/rss/creativeCommonsRssModule.html">
    <channel>
        <title><![CDATA[Stories by Robert P. Levy on Medium]]></title>
        <description><![CDATA[Stories by Robert P. Levy on Medium]]></description>
        <link>https://medium.com/@rplevy?source=rss-3c03111694de------2</link>
        <image>
            <url>https://cdn-images-1.medium.com/proxy/1*TGH72Nnw24QL3iV9IOm4VA.png</url>
            <title>Stories by Robert P. Levy on Medium</title>
            <link>https://medium.com/@rplevy?source=rss-3c03111694de------2</link>
        </image>
        <generator>Medium</generator>
        <lastBuildDate>Sat, 21 Nov 2026 11:10:56 GMT</lastBuildDate>
        <atom:link href="https://medium.com/@rplevy/feed" rel="self" type="application/rss+xml"/>
        <webMaster><![CDATA[yourfriends@medium.com]]></webMaster>
        <atom:link href="http://medium.superfeedr.com/" rel="hub"/>
        <item>
            <title><![CDATA[Propositions Are Not Types: Naturalizing Information Content in Computing]]></title>
            <link>https://medium.com/@rplevy/propositions-are-not-types-naturalizing-information-content-in-computing-e51b94ad6674?source=rss-3c03111694de------2</link>
            <guid isPermaLink="false">https://medium.com/p/e51b94ad6674</guid>
            <category><![CDATA[embodied-cognition]]></category>
            <category><![CDATA[cybernetics]]></category>
            <category><![CDATA[augmented-intelligence]]></category>
            <category><![CDATA[artificial-intelligence]]></category>
            <category><![CDATA[functional-programming]]></category>
            <dc:creator><![CDATA[Robert P. Levy]]></dc:creator>
            <pubDate>Mon, 31 Dec 2018 19:36:02 GMT</pubDate>
            <atom:updated>2018-12-31T19:36:02.878Z</atom:updated>
            <content:encoded><![CDATA[<p>Overview: “the hard problem of content” (Hutto &amp; Myin 2013) is an argument that information as it exists in physical systems is only covariance information, which does not inherently have properties such as truth, reference, and implication required for it to work as propositional content. That means formal propositions implemented in computing systems can’t actually be carriers of content as information-processing theorists had hoped. But in order for software engineering to mature into reliable practice, we will need computing systems to participate in propositional activities on our behalf. Following Hutto &amp; Myin’s solution that real-world content depends on shared scaffolded practices that augment basic agents, we explore socially situated computing as an approach to naturalizing information content in computing.</p><h3>Part I: Propositions are Not Types</h3><p>Over the last decade or so a research finding that was previously heard mostly in university math and computer science departments has become a popular slogan in applied statically typed functional programming circles. The message that inspires this movement: “propositions are types! programs are proofs!” (Baez &amp; Stay 2009) It’s not a new idea that leading figures such as Philip Wadler and Simon Peyton Jones promote, it is a well-established finding that unfolded over a period from 1934 to 1969 in the work of mathematicians Haskell Curry and William Alvin Howard. From a purely mathematical standpoint there is nothing false about the Curry-Howard Correspondence as it is called. It is indeed useful, as it allows computer programs to be analyzed using the tools of formal logic, everything from proof assistants to compile-time type-checkers in general-purpose programming languages. Pretty neat.</p><p>Where we start to run into problems is that the emboldened typed functional language communities increasingly advocate for a radical perspective on logic and information processing that is reminiscent of the symbolic processing era of AI. Their hope or expectation is that the programs-as-proofs approach will scale to messier, more complex real-world applications of formal methods, transforming computing from the status of unreliable and irresponsible craft to robust engineering discipline. The goal of being able to declare what our systems must abide by and hold to is laudable and I share it, but the means to that goal brush over some thorny problems.</p><h3>Lessons from Artificial Intelligence and Cognitive Science</h3><p>Computationalism or “GOFAI” did not exactly die when it failed spectacularly in the mid-80s, but in the time since the AI winter there has been progress on understanding why such approaches are bound to fail. As new research on pre-AI cybernetics-era techniques such as neural networks began to show improvements, “symbol grounding” was identified as an overlooked problem with formal propositional systems. The symbol grounding problem is that symbols’ interpretations are not intrinsic to formal systems, but rather depend on users of a system to assign their meanings. Consider how any property of the world is encoded as data. Symbol grounding problems arise due to encoding of properties and events in ways that arbitrarily coincide with human interpretation, but do not intrinsically preserve lawful relationships to aspects of the world. Today computationalists in AI seem to be excited about combining the more grounded behaviors of neural networks with symbolic processing ideas, but symbol grounding remains a difficult problem for the information-processing paradigm generally.</p><h3>The Hard Problem of Content</h3><p>But there is an even deeper problem with information processing approaches to information content. Fred Dretske, the thinker perhaps most emblematic of the original computationalism, proposed that a “cognitive” system (meaning a system acting on information content) is one that converts analog information into a digital representation, and performs manipulations on that digital information, allowing it to perform actions targeting properties of its environment. Dretske’s notion of information is that data indicates or carries information about some property of the world if the two vary lawfully enough. Recently Dretske’s theory of content has met with a strong challenge from cognitive science philosophers Daniel Hutto and Erik Myin, that such accounts of content have no grounds on which to claim that lawfully covarying information “says” or “represents” anything about the world. This observation which they refer to as the “hard problem of content” or the “covariance-is-not-content principle” is that systems acting on covariance information, while acting on information, do not constitute content-bearing systems, because to bear content is to embody claims about how things stand, when in fact they merely embody capacities to affect the world.</p><p>So even a well-grounded symbol is not intrinsically about anything, it cannot be about something, it only lawfully varies in relation to some property of the world. The consequence of this for the design of computing systems is that we shouldn’t expect to find a way for computing systems much less biological agents to work with self-contained information-as-content when all we’ve given them is information-as-covariance lacking any way of participating in its larger social/pragmatic context. What we should look for instead are ways of scaffolding propositional interactions in effective world-involving ways. As Wittgenstein surmised after a career spent grappling with the basic nature of propositions, propositions themselves do not contain information about how things stand, rather they are like moves in games where the object is to know how things stand in the world.</p><h3>Part II: Naturalizing Information Content</h3><p>We’ve talked about covariance information, and how it is not the same thing as information content, but how can we naturalize content? In order to answer that, it is necessary to take a step back from a narrow focus on abstract information, and look at human-computer interaction, which is a special case of human-environment interaction. We will touch briefly on how basic agents use information effectively and produce it expressively, and then discuss how convention-based access to such information (via cooperative scaffolding of motivated attentional decisions) is leveraged in claim-making games– and how socially situated computing systems can participate in these “games”, reaping the benefits of participating in authentic information content.</p><h3>Lawful and Conventional Access to Ecological Information</h3><p>Embodied cognitive science research has shown that covariance information, while content-free and non-representational, can go a long way in explaining a host of complex cognitive behaviors. The most scientifically mature effort in embodied cognitive science is ecological psychology, founded by J. J. Gibson in the 1950s, 60s, and 70s. Ecological psychology has made as much progress as it has on naturalizing phenomenology by discovering that organisms acquire “ecological information”, information supporting perception/action, in the form of learning to exploit what are called “affordances” or physical properties of the environment that reliably specify directly realizable actions/outcomes to organisms. Ecological information is law-based covariance information embodied by organisms guided by affordances in their prospective control and navigation of environments.</p><p>But ecological information supporting the control of action can be accessed in a secondary way, by means of convention, as Sabrina Golonka published in 2015. Research has found that only law-based information can support direct perception, but conventions such as signs and gestures can be used to select or steer attention to a primary target of direct perception, and as such constitute a mode of access to ecological information. Importantly, such expressive conventions do not constitute content and representation. As José Medina (2013) puts it, basic convention-based expression “should not be understood on the Gricean model of conventional signs, that is, as involving or requiring fully formed communicative intentions and internal representations. Expressive behavior is not self-reflective intentional-referential behavior among rational agents who are representing each other’s minds and their contents.” Rather, conventional expression is a mechanism of directing attention.</p><h3>Joint Attention, Scaffolding, and Claim-making</h3><p>Conventionality does not in itself grant such sophisticated uses as representing content, rather it serves as the foundation for them. As Hutto &amp; Myin (2017) write:</p><blockquote><em>“content only arises when special sorts of sociocultural norms are in place. The norms in question depend on the development, maintenance, and stabilization of practices involving the use of public symbol systems through which the biologically inherited cognitive capacities can be scaffolded in particular ways. The practices in question are claim-making practices– and they are special because they require participants not only to respond to things but to do so by </em>representing them as being thus and so<em> independently of what might be said about them.” (italics theirs)</em></blockquote><blockquote><em>…</em></blockquote><blockquote><em>“Getting things wrong in a truly representational sense is not just a matter of being literally misguided in the way purely biological entities and creatures can be. It involves being subject to the censure of others– not just in the sense of being in or out of line with what is acceptable or not for some community, but being able to get things wrong in a game in which it is at least possible to be right according to how things are anyway. Only those in a position to play this sort of game can be said to have content-involving thoughts and speech.”</em></blockquote><p>So then what does this socially situated “scaffolding” of basic direct and conventional access to ecological information require? The most empirically compelling answer to this question is to be found in Michael Tomasello’s research on primates and child development. Tomasello has found that primates use gestural conventions classified into two categories: “attention-movement” gestures, to get another agent to do a particular thing, and “attention-getter” gestures that call attention of another agent to something that they’ll respond to in some way. Humans in contrast (at around 9 months of age) develop a more powerful vocabulary of social attention control devices. Humans go through three stages of learning how to control attention. The first is “sharing” what they are attending to, not unlike the attention-getter techniques of apes. The second is “following into”, as in attending to what another agent is attending to. The third developmental stage is the skill of “directing” others to attend. Directing is the most impressive skill out of the three because unlike the attention-mover gestures of the great apes, directing attention occurs relative to a followed-into shared context of attending. It is worth noting that these social skills appear several years earlier in child development than the “theory of mind” skills. Because of their early development and marked divergence from other primates’ functionally similar abilities, Tomasello theorizes that they constitute an innate and evolved “infrastructure of shared intentionality” supporting cooperative communication that paves the way for complex tools of cooperation such as spoken language.</p><p>The development of joint attentional skills was a defining moment in becoming human as we know it. It made it possible for humans to construct attentional tools (paintings, glyphs, models etc) that augmented their gestural scaffolding of attention. Such scaffolding devices included the development of language itself, in which verbal constructions are literally used as tools. With joint attention, and its augmentation by scaffolding, we approach the aforementioned Gricean account of communication as prosocial, cooperative activity. This gives us the necessary ingredients for the social construction of claim-making scenarios, such that one might play the game and be successful or fail, with a given propositional move, at achieving socially defined objectives in a shared environment.</p><h3>Scaffolding, Constructions, and Conceptual Metaphor</h3><p>I would like to take a moment here to revisit the insufficiency of the formal propositional account of information content that is the focus of Part 1 of this article. There is a potential objection to our positive neo-Wittgensteinian account (that propositions are less like pictures or containers, and more like moves played in games), the objection being that any such game moves can only be smaller fragments of world models, and that the problem has simply shifted to a finer grain. We have already seen one way in which that is not the case, that is in pre-linguistic deictic social skills of joint attentional engagement. But that is not yet true claim-making, so what of the scenario of mature claim-making contexts? The fact is that the objection ignores that language even in its simplest cases does not consist in formal world modeling but in guiding and motivating flows of attention in a collaborative process of narrative sense-making. For consistency, I will assume the usage-based model of language constructions (Tomasello 2003, Goldberg 2006) as tools (which I also happen believe is true, given its elegance and empirical track record.) Humans are at base engaged in joint attention for the purpose of cooperating on activities, and language affords powerful leverage in those processes. As an example, around 2–3 years of age children begin to pick up on identification and possession constructions like “it is X”, “that is X”, and “that’s my X”. Using these constructions is to participate in engaged processes, supporting them by calling attention to something someone would presumably want to know. Further, it has long been well-established scientifically that conceptual metaphor (Ortony 1993), the practice of adapting familiar schemas from basic-level perception to make sense of or define more abstract or complex ideas. Conceptual metaphor is a an application of joint attentional scaffolding that human languages get a great deal of mileage out of. Again, this is a way of resolving sense-making to flows of motivated attention. So it is not a deflective move to take the pragmatic turn on propositionality, rather it is to embrace the reality that cooperative attentional scaffolding is the basis of sense-making and communication, and that the claim-making scenario (where what is expressed may win or lose, be correct or incorrect) is no exception.</p><h3>Socially Situated Programming</h3><p>The picture we have arrived at is of human culture as a large-scale stigmergic content ecosystem, consisting in human social transaction within contexts of content generation, testing, and upkeep. Individuals from a young age are confronted with the need to make myriad passive and active attentional decisions that become in some large part what is unique about the life experience, perspectives, and directions of any particular person. As social creatures we enter a world that offers us a wealth of pre-existing tools for directing attention in useful ways, and games or systems establishing utility of actions. Despite all of culture’s complexity, there is always one thing happening that makes content possible: scaffolded processes of attending in game-like social contexts.</p><p>For designers of interactive computing environments, including programming languages and other kinds of systems in need of open-ended declarative expression, the trillion-dollar question is: how can machines participate meaningfully and effectively in content? I will present two answers to this, one too general for it to be immediately obvious how it might be put to use, and one too specific to imply any sort of claim about the space of other possible solutions– just exploring one approach in some depth. It is left to the reader to experiment with other possibilities in this space.</p><p>The very broad answer can be summarized as “socially situated computing”. To define this I will start with defining the more encompassing “situated computing”. The idea of situated computing or situated programming is that computing is embedded more directly into situations than it traditionally has been, putting systems into shorter real-time feedback loops with events of interest to its users. Rich Hickey has emphasized the importance of this embeddedness and feedback for creating reliably effective software systems, whereas others such as Jelle van Dijk, M Eifler, and Bret Victor have focused on the implications of immersive technologies on embodied engagement with direct environments as we experiment with a plethora of new kinds of devices, sensors, and instruments. Others such as William J. Clancey and Rodney Brooks have focused on the importance of environment embeddedness for the intelligence of artificial agents. The common theme in all of these approaches is a reorientation of computing to be more ubiquitously agent-centered, context-sensitive, and feedback-oriented.</p><p>Phoebe Sengers (1996) coined the term “socially situated AI” to refer to approaches to AI that are not only aware of the agent’s relationship to its physical environment, but also its social environment. Expanding the scope of this idea a bit, social-world-involving software augmentation of experience is what I am calling socially situated computing. It is my contention that from an HCI perspective, an AI perspective, and a generally informatic perspective, embedding of computing into the contexts where content is maintained is a requirement in order to make systems content-aware.</p><h3>Narrative Process Scaffolding</h3><p>To answer the question posed earlier “how can machines participate meaningfully and effectively in content” with the broad brush of “socially situated computing” is appropriately non-committal given the current nascent state of the art, but at the same time it is unsatisfying because it doesn’t get into specifically <em>how</em> one might go about involving computing in contexts where content is maintained. As I see it, this problem has two sides to it: a causal decision science problem (Pearl 2000, 2018) and a human-computer interaction problem: on the human side, expression of causal motivation of attention, and on the machine side, causal inference by agents selecting actions that contribute helpfully within these contexts. Following Tomasello and Carpenter’s work on joint attention, the attentional skills of sharing, following-into, and directing are intuitively accessible to 9-to-12-month old babies. This research suggests that the skillful navigation of contexts of attending constitutes our ability to express and understand intentions. In other words, intending is decision-making about attending (and nothing else). If you know <strong>why</strong> an agent is attending, then you know what they are intending (limited to that specific context, of course). “Why attending” and “what intending” are different ways of expressing the same thing.</p><p>By establishing a protocol codifying joint attentional decisions into a set of simple gestures, we can make use of the basic joint attentional skills that we’ve leveraged for sharing motivation of attention since early childhood, to specify learnable contexts that software agents can participate in, extending our agency as users. To be a little more specific, there seem to be three kinds of attentional decisions: whether or not to enter into a context or center of attending, whether or not to attend to some other center in the context of the current one, and given the presence of an option to attend to another center, whether or not to exit. The causal questions that can be expressed using these primitives are respectively “does attending to or refraining from attending to this center cause this outcome”, “does attending to this other center influence positively or negatively the outcome of the present center”, and “does the availability of this other center indicates the outcome has been reached or is being maintained”. It’s important to note here that these causal models are propositional in our human-to-human description of them in talking about the approach, but in use they are modeling the causal properties of non-propositional capacities of expression. It is through the use of these capacities in explicitly claim-making processes that actual machine-accessible content emerges. I have called the approach of using <em>in situ</em> gestures to define causal models that scaffold intentional agent behaviors “narrative process scaffolding” (Levy 2018) because motivating joint attention in this way is theorized to be fundamental mechanism of narrative sense-making generally, including processes of claim-making that underpin propositional content.</p><p>The narrow prescription given here for “narrative process scaffolding” is still speculative and untested, but if it is developed into a robust practical computing paradigm, it is preferable over the formal propositional program that we have stated strong reasons to believe can’t work. NPS in contrast is guided by findings on how information content does appear to work. Further, if NPS turns out to be somehow irrecoverably flawed in its approach, it should be kept in mind that this is just one idea of how to design socially situated computing systems. The space is a wide open frontier for creative and scientific exploration.</p><h3>From Metaverse to Noosphere: A Third Cybernetics?</h3><p>Today we see the early signs of a third wave of cybernetics emerging. The first cybernetics tackled teleological mechanisms as physical control systems. Many of the most successful “AI” techniques today (neural networks, reinforcement learning, agents as dynamic control &amp; feedback systems) began in that movement, which is ironic because AI and its attendant “cognitive revolution” arose in opposition to cybernetics. The second cybernetics sought an understanding of second-order or self-organizing autonomous systems. Such efforts eventually became part of 4E Cognitive Science: embodied, embedded (situated), enactive, and extended (scaffolded). The potential ramifications of socially situated computing and related trends suggest the third cybernetics will turn a microscope, and ultimately a telescope, on the 3rd-order processes by which autonomous self-organizing agents engage in joint attentional cooperation, and participate in claim-making processes constituting the narrative construction of the content of all of culture and conceptual thought. The third-order cybernetics is the one that would expose the workings of culture and personality to tractable computation.</p><p>Situated and socially situated computing is on the rise. There is renewed interest in and success of agent-based and agent-centered approaches in computing. Giant companies pour research money into nebulous efforts toward “cognitive computing” as a goal of agent-based computing approaches such as robotic process automation. There is new life being breathed into augmenting human agency, human-machine symbiosis, and intelligence amplification. The promise of “the metaverse” (the situated internet) is being cultivated in emerging web standards for augmented reality on the open internet. Self-sovereign identity, user ownership of data, and attentional sovereignty make the prospect of socially situated content-involving metaverse, an “electronic noosphere” more ethical and humane. Importantly securing sovereignty of identity and attention provides the base of necessary confidence required for participation in systems that inherently demand great depth and intimacy of integration of computing into personal and social experiences. Such deep integration of computing into human sense-making is likely to radically alter the human cognitive niche in ways that we as future individuals and networks will be in control of, but which appear impossible to form any detailed expectations about at present without actually beginning to undergo it.</p><h3>References</h3><ul><li>“Philosophical Investigations” Wittgenstein 1953 <a href="https://static1.squarespace.com/static/54889e73e4b0a2c1f9891289/t/564b61a4e4b04eca59c4d232/1447780772744/Ludwig.Wittgenstein.-.Philosophical.Investigations.pdf">https://static1.squarespace.com/static/54889e73e4b0a2c1f9891289/t/564b61a4e4b04eca59c4d232/1447780772744/Ludwig.Wittgenstein.-.Philosophical.Investigations.pdf</a></li><li>“The Ecological Approach to Visual Perception” Gibson 1979 <a href="http://b-ok.cc/book/864226/e0dd92">http://b-ok.cc/book/864226/e0dd92</a></li><li>“Knowledge and the Flow of Information” Dretske 1981 <a href="http://b-ok.cc/book/819455/a1f712">http://b-ok.cc/book/819455/a1f712</a></li><li>“Intelligence without representation” Brooks 1991 <a href="https://people.csail.mit.edu/brooks/papers/representation.pdf">https://people.csail.mit.edu/brooks/papers/representation.pdf</a></li><li>“Metaphor and Thought” Ortony 1993 <a href="http://b-ok.cc/book/767226/2778f8">http://b-ok.cc/book/767226/2778f8</a></li><li>“Socially Situated AI: What It Means and Why It Matters” Sengers 1996 <a href="https://pdfs.semanticscholar.org/1414/7b6006180a87c139a29f0cc1584b8437e915.pdf">https://pdfs.semanticscholar.org/1414/7b6006180a87c139a29f0cc1584b8437e915.pdf</a></li><li>“Computation and Human Experience” Agre 1997 <a href="http://b-ok.cc/book/689039/b2182d">http://b-ok.cc/book/689039/b2182d</a></li><li>“Situated Cognition: On Human Knowledge and Computer Representations” Clancey 1997 <a href="http://b-ok.cc/book/859418/09e510">http://b-ok.cc/book/859418/09e510</a></li><li>“Social cognition, joint attention, and communicative competence from 9 to 15 months of age.” Carpenter, Nagell, Tomasello 1998 <a href="http://booksc.xyz/book/15451904/9e3c39">http://booksc.xyz/book/15451904/9e3c39</a></li><li>“Causality” Pearl 2000 <a href="http://b-ok.cc/book/907051/f3c24a">http://b-ok.cc/book/907051/f3c24a</a></li><li>“Constructing a Language: A Usage-Based Theory of Language Acquisition” Tomasello 2003 <a href="http://b-ok.cc/book/690493/92df60">http://b-ok.cc/book/690493/92df60</a></li><li>“Constructions at Work: The Nature of Generalization in Language” Goldberg 2006 <a href="http://b-ok.cc/book/1210861/e29e0d">http://b-ok.cc/book/1210861/e29e0d</a></li><li>“Origins of Human Communication” Tomasello 2008 <a href="http://b-ok.cc/book/541274/39859f">http://b-ok.cc/book/541274/39859f</a></li><li>“Physics, Topology, Logic and Computation: A Rosetta Stone” Baez &amp; Stay 2009 <a href="http://math.ucr.edu/home/baez/rosetta.pdf">http://math.ucr.edu/home/baez/rosetta.pdf</a></li><li>“Radicalizing Enactivism” Hutto &amp; Myin 2013 <a href="http://b-ok.cc/book/2554656/1b7ea8">http://b-ok.cc/book/2554656/1b7ea8</a></li><li>‘An Enactivist Approach to the Imagination: Embodied Enactments and “Fictional Emotions”’ José Medina 2013 <a href="https://www.jstor.org/stable/24475354?seq=1#page_scan_tab_contents">https://www.jstor.org/stable/24475354?seq=1#page_scan_tab_contents</a></li><li>“Laws and Conventions in Language-Related Behaviors” Golonka 2015 <a href="http://booksc.xyz/book/50082310/1e8631">http://booksc.xyz/book/50082310/1e8631</a></li><li>“Strange Tools: Art and Human Nature” Noë 2015 <a href="http://b-ok.cc/book/2640649/b1b44d">http://b-ok.cc/book/2640649/b1b44d</a></li><li>“You’re Doing Mixed Reality Wrong” Eifler 2017 <a href="https://medium.com/@blinkpop/youre-doing-mixed-reality-wrong-d32aa54ae8af">https://medium.com/@blinkpop/youre-doing-mixed-reality-wrong-d32aa54ae8af</a></li><li>“Effective Programs” Hickey 2017 <a href="https://github.com/matthiasn/talk-transcripts/blob/master/Hickey_Rich/EffectivePrograms.md">https://github.com/matthiasn/talk-transcripts/blob/master/Hickey_Rich/EffectivePrograms.md</a></li><li>“Evolving Enactivism” Hutto &amp; Myin 2017 <a href="http://b-ok.cc/book/2947353/09d772">http://b-ok.cc/book/2947353/09d772</a></li><li>“Steps to an Ecology of Bicycles for the Mind: A Situated Programming Manifesto” Levy 2018 <a href="http://senters.info/situated-programming">http://senters.info/situated-programming</a></li><li>“The Book of Why” Pearl 2018 <a href="http://b-ok.cc/book/3559945/bd1c4f">http://b-ok.cc/book/3559945/bd1c4f</a></li><li>“Designing for Embodied Being-in-the-World: A Critical Analysis of the Concept of Embodiment in the Design of Hybrids” van Dijk 2018 <a href="http://www.mdpi.com/2414-4088/2/1/7/pdf">http://www.mdpi.com/2414-4088/2/1/7/pdf</a></li><li>“Reflections on Dynamicland” Wittke 2018 <a href="https://medium.com/@wittkensis/reflections-on-dynamicland-65158b06196">https://medium.com/@wittkensis/reflections-on-dynamicland-65158b06196</a></li></ul><p>This article is maintained at <a href="https://github.com/senters/pant/blob/master/pant.org">https://github.com/senters/pant/blob/master/pant.org</a></p><p>Visit <a href="http://senters.info/">http://senters.info</a> to learn more about the senters project of implementing narrative process scaffolding.</p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=e51b94ad6674" width="1" height="1" alt="">]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[Steps to an Ecology of Bicycles for the Mind]]></title>
            <link>https://medium.com/@rplevy/steps-to-an-ecology-of-bicycles-for-the-mind-95fa649ef422?source=rss-3c03111694de------2</link>
            <guid isPermaLink="false">https://medium.com/p/95fa649ef422</guid>
            <category><![CDATA[augmented-reality]]></category>
            <category><![CDATA[ecological-psychology]]></category>
            <category><![CDATA[joint-attention]]></category>
            <category><![CDATA[intelligence-augmentation]]></category>
            <category><![CDATA[situated-programming]]></category>
            <dc:creator><![CDATA[Robert P. Levy]]></dc:creator>
            <pubDate>Sun, 13 May 2018 21:19:28 GMT</pubDate>
            <atom:updated>2018-05-13T21:19:28.431Z</atom:updated>
            <content:encoded><![CDATA[<p><strong>A Situated Programming Manifesto</strong></p><p>[Note: the print-friendly zine-style packaging of this text can be found at <a href="http://senters.info/situated-programming/">http://senters.info/situated-programming/</a>]</p><h3>PREFACE</h3><p>In the folklore of computing, the turn of phrase “bicycle for the mind”¹ characterizes the aims of intelligence amplification (IA) efforts by way of reference to the well-established scientific fact that bicycles come astonishingly close to making total use of a person’s self-generated mechanical power. A bicycle is the most efficient vehicle to harness a human’s own power toward locomotive objectives.² Can computing technology similarly be used to maximally leverage human powers of perception/action and narrative sense-making?</p><p>The metaphor raises an obvious question: why not an electric bicycle, or a self-steering sailboat, or a telepresence drone — for the mind, for that matter? Power tools of thought are ultimately what we want, but there is risk in putting them first. The bicycle model of amplifying human powers highlights the challenge of how to extend agency with technologies closely fit to core human competencies and needs. Skipping this necessary first step seems to lead down a path of not augmenting human capacities, but instead supplanting them in ways that tend to integrate poorly with the actual interests and needs of users. It leads to inscrutable oracles that trade off accountability and expertise for short-sighted convenience.</p><p>We do not yet have anything quite analogous to bicycles for the full power of the embodied human mind. But the history of human life has been punctuated by the development of new practices and tools for scaffolding and extending personhood. I suppose we could think of some of these innovations as wheelbarrows or shovels for the mind, augmenting capacities to participate in the world in specific ways. To measure up to the standards of the analogy from human mechanical power, a bicycle for the mind must extend core human competencies. What skills account for our unique cognitive ecological niche?</p><p>A clue is found in our collaborative graphical practices. The very earliest archaeological evidence of modern humans is littered with tools designed to cooperatively guide attention. In more recent human epochs, the printing press, mass media, and the internet have expanded the reach of graphical practices. Scarcity of attentional resources is a limiting factor, one historically addressed institutionally. However, we have reached a point where institutions are a bottleneck too. In order for networked individuals to viably maintain situation awareness at scale, it is necessary for software to participate in the content of information, not merely its mechanical transport and processing. Computing tools must be capable of following and augmenting our intentional processes. In other words we need bicycles for the mind.</p><p><strong>The thesis put forward in this document, following research in ecological perception and joint attention, is that direct attending is reorganized cooperatively through use of conventions for selecting, composing, and realizing feedback loops of attending. The prescription for bicycles for the mind that follows from this is to codify the scaffolding of attending for machine followability, using immersive technology, to augment situation awareness and sense-making.</strong></p><p>I call this work a “manifesto” because I don’t consider it an idle theoretical exercise, but instead as advocacy for certain approaches and courses of action that are motivated by real problems in society, technology, and science. Alan Kay once remarked that “the best way to predict to future is to invent it”. Likewise the best way to make sense of history is to excavate from it the most sound basis for the direction one is inventing toward. I survey three major trends tracing all the way back to the birth of modern human primates: attending, working, and organizing. Starting with the origins of joint attentional scaffolding and following a thread through changes that have impacted the scale of attention, work, and social organization, I draft a roadmap to an ecology of bicycles for the mind. I have kept references to authors, works, and prior art in the endnotes apart from the main text to simplify presentation.</p><h3>AGENTS AND THEIR ENVIRONMENTS</h3><p>Two of the oldest questions in philosophy and science are “what is life?” and “what is mind?”. Only a mere 150 years ago¹ did it begin to dawn on pragmatic thinkers that these two questions are actually one question: “what is an agent?”. To ask what an agent is is to ask what the difference is between inert mechanisms and living, feeling, self-steering mechanisms. So the question naturally becomes one of feedback and control mechanisms. Agents are systems that act to control parameters of environments that matter to them. To speak of an “environment of” or “mattering to” a system requires a physical explanation, and that explanation is given by natural selection. Mattering is a consequence of evolution because physico-chemical arrangements are either ephemeral or they persist through targeted manufacture, maintenance, acquisition of resources, and so on. To persist requires control of factors bearing on persistence, as biases or preferences. In other words, those things matter to the agent — whether the agent knows it or not.</p><p>An agent, minimally defined, is a system that exercises control over factors ultimately tied to its persistence as an autonomous entity over time. Factors that may be required for the unity of the system to endure but are less directly controllable by the agent are called the environment of an agent. It is important to note that up until here, when we speak of mattering to an agent it is meant strictly in terms of outcomes that can be observed in the functioning of the system by an outside observer. What is called “perception” is an agent’s active sensitivity to ecological features called affordances that bear upon control objectives. To say affordances matter to an agent, is to say that the real outcomes matter (both in terms of learning effective habits and in terms of reproduction/survival viability), and to remark on the first-hand phenomenology of proactively leveraging ecological information.</p><p>Agents’ end-directed prospective control can be described in terms of three predicaments or stages in processes at which different factors bear upon action. Evidence for these predicaments derives from the pragmatics of gesture in joint attention, explored further in the next chapter.² Their efficacy in communication suggests they constitute the ontology of agency.³</p><ol><li>ATTENDING: An agent is situated in environments that affords many opportunities and threats demanding awareness and readiness to potentially engage promising ends and mitigate perilous eventualities. Any engagement has opportunity costs, so an agent must budget attention⁴ wisely. <strong>The predicament of attending is that of classifying events as indicative of means to valued ends, in deciding to engage or refrain.</strong></li><li>SUPPORTING: Support, in the sense of contribution of work or effort, refers to relationships that hold between a presently engaged process and one or more potential or ongoing process. <strong>The predicament for the agent is to what extent does the work of a secondary engagement contribute to (or hinder) the realization of the present enagement’s objective, in deciding to include in scope or not.</strong> The phenomenon of concentration or focus is also expressed by optimally excluding/including supporting or subordinate engagement. The idea of focus is that only those supporting processes that contribute to realizing valued ends should be involved, and others should be excluded.</li><li>REALIZING: The third predicament that agents contend with is that of organizing final causes. Final causes are those that motivate the second predicament above and indirectly, the first as well. <strong>The predicament is that of defining ends, in deciding to continue or exit. What defines the unique character or meaning of realized ends is the specific set of options that are available to exit to.</strong> In general, positions that offer more options (giving the agent more autonomy, rather than being at the mercy of circumstance) appear to be favorable, all other things being equal.⁵</li></ol><h3>JOINT ATTENTION: THE HUMAN NICHE</h3><p>Humans are a unique case among natural agents. What distinguishes the “sapience” of humans from the more widespread “sentience” of basic agents is its narrative character. The source of our narrative abilities is the topic of this chapter. Narrative presentation or representation is not deep or intrinsic to intelligence by any means. The rich variety of sentient intelligence in nature manifests as purely pragmatic capacities of agents to steer events toward ends that matter — solely in terms of direct bearing on positioning or readiness, often to the mutual benefit of agents.</p><p>Narrative agency is wholly due to a set of evolved capacities constituting a cooperative infrastructure of communication that began with gesture.¹ The basic gestural functions call attention to predicaments of agency: attending, supporting, and realizing processes. These skills allow us to collaboratively develop ways of seeing and knowing, by intervening to fine-tune the selection, composition, and motivation of decisions of what to attend to and why. While other species do engage in joint attention², only humans attend to a shared focus and coordinate motivation for directing and refining that focus.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1004/1*nOjQF1PjycuvB1ObqLXtKA.png" /></figure><p>Joint attention is important for multiple reasons. One reason is that it is directly responsible for human intentionally organized tools and practices. This is because when agents externally coordinate events that bear upon attending, supporting, and realizing, they are defining and shaping shared processes of narratively reorganized agency.³ Another reason, a consequence of the first, is the development of intersubjectivity: the practices and tools that arise in service of joint attention become a rich repository of scaffolding that constitutes tacit and explicit common ground supporting the many varieties of social interactions. The accumulation of scaffolding offers potentially endless options for agents to extend the reach of their world-involving activities in socially sharable ways. Somewhat ironically, social practices establishing content made possible the development of autonomous personal “selves” as nexuses of life narrative, identity, and “mental” contents.</p><p>The earliest stages of human societies mark the beginnings of a history of tools and practices (such as paintings, masks, signs, speaking, etc) for scaffolding joint attention. Scaffolding is still by and large a highly localized activity between small groups of people, though the coordination of local activities has been significantly broadened by technologies of writing, the printing press, mass media, and the internet. Disseminable graphical practices such as writing and diagramming exemplify how <em>extensive⁴</em> sapient agents are: the reach and influence of involvement in the world is amplified by distributed scaffolding practices. Bicycles for the mind aim to tighten or make more immediate these wide feedback loops so as to scale awareness, sense-making, and organizational practices.</p><h3>THE SCALABILITY OF ATTENTION, WORK, AND SOCIAL ORGANIZATION</h3><p>In the discussion above we’ve developed an account of narrative agents whereby intervening in other agents’ attending to three basic predicaments of control (by means of gesture and scaffolding), agents establish sophisticated narrative sense-making skills and practices. The human niche in all of its cultural and psychological diversity has been built in this framework of pro-social sensitivity to other agents’ states of attending to, supporting, and realizing objectives. From these control predicaments and their derived communicative motives and skills emerge corresponding cultural phenomena.</p><ul><li>ATTENDING Any social unit, however temporary or time-extended, and however large or small, is confronted with the need for coordination on what merits attention. Agents develop a shared awareness or common involvement in making sense of events as bearing on shared concerns.</li><li>WORK (SUPPORTING) Any social unit is confronted with the need for ways of having effects on the world together. Tools and machines have greatly expanded our capacities to work toward objectives. Technologies of attention, such as diagramming and notation, serve to affect agents’ processes of realizing specific ends.</li><li>SOCIAL ORGANIZATION (REALIZING) Any social unit is confronted with the need to coordinate on final causes that constitute “what we want to happen”. This can include anything from a preference to avoid places with cold temperatures in an outing together, to political causes, legal justice rulings, or the incentive structure of a game.</li></ul><p><strong>Hunter gatherer/tribal society.</strong> All evidence suggests that the primary original driver of increasingly advanced cooperation amplified by joint attention in proto-human primates was the stag hunt. Big game hunting required more than one person’s help, and the better they coordinated, the better results they got on average. However early human societies did not gather in groups much larger than other primates did, because of the limitations of keeping track of one another and coordinating collective efforts. Humans developed intentional tools for practical purposes (like spears), and tools and practices of attention such as gesture, drama, ritual, painting, symbols, and an explosion of cultural artifacts as they began to bootstrap their capacities further, scaffolded in this way. It was in this period that people began to tell stories and make sense of nature in the best ways they could find to influence their shared fate positively.</p><p><strong>Agricultural/hierarchical society.</strong> The next major shift in human social organization was driven by agriculture. Permanent settlements inhabited by unprecedented numbers of people led to increasingly hierarchical social organization to manage bottlenecks of attention, work, and governance at scale. In this period the hallmarks of civilization, coins for example (out of the collectible and gift exchange practices of forager societies) arose to address limitations of personal and collective situation awareness that rendered ineffective older established ways of sharing in common resource pools. Mechanical devices and managerial hierarchies were developed to address the scale of work. Heavy-handed monarchies and religious institutions addressed problems of coordinating social values at scale.</p><p><strong>Information/network society.</strong> In the present era we are at the beginning of another major shift in the scale of society, from a hierarchical civilization to a decentralized network society. Near the end of hierarchical civilization, the human project of industrialization culminated in a generalization of efficient work in the form of information theory, control theory, and the invention of the digital computer. Networked computers have globalized the coordination of mechanical work, enabling ad hoc, easily affordable, planetary-scale machinery. Just as old methods failed us when we formed large settlements, the new global and interplanetary scale of human interaction necessitates decentralized social organization. Beyond developing robust distributed architecture discipline, what is needed most is scaling shared situation awareness and organization.</p><h3>SITUATED PROGRAMMING</h3><p>At the traumatic height of the 20th century, cybernetics emerged as a bold scientific, technological, and societal program that emphasized the engineering of agent mechanisms. Cybernetics was instrumental in the invention and development of computing technology, which it cast as a tool for designing and understanding systems organized by mechanisms resembling the end-directed processes of natural agents. The first high-level programming languages reflected the machine and human sides of this project respectively, with languages like Fortran focusing on efficient use of machine resources, and languages like Lisp and Smalltalk focused on amplifying human expressive power. As time went on and resources grew more abundant, Expressively powerful languages became more practical. Parallel to these developments, user bases were expanding, but at the cost of a growing divide in power between “programmers” and “users”. Ever since that trend started, there has been a tension in computing between programming being disconnected with situated use, and use lacking the expressive power of programming. Consequently there have been many efforts to make programming more accessible to wider audiences, and/or to make situated use as powerful as programming. The setbacks such efforts have faced have been both technical (fundamental challenges in interface design, knowledge representation, etc.) and cultural (respectability politics in academia and industry, economic factors).</p><p>For a few different reasons, the tide seems close to turning. Mobile and wearable computing is increasingly ubiquitous and we are only a couple of years away from affordable consumer augmented reality hardware. That means computing is more frequently applied to activities other than just those ones that are best done sitting at a desk. Some of the original cybernetics ideas, namely neural networks and agent-based learning have become wildly successful, and there is a growing interest in ways of integrating machine learning techniques with data-centric approaches to programming. Efforts in industry such as robotic process automation are aptly proposing agents as an effective user interface model to give users more expressive power. The architecture pattern of event-driven dataflow is gaining in popularity at the same time that decentralized data provenance infrastructure is maturing. Data provenance is the cornerstone of decentralized network society, because as computing scales the labor dimension of collaborative agency, it becomes necessary to scale social accountability with respect to effects and motivations of work. Having auditable records is essential to assessing and correcting faults, and for learning what works and why, in repeatable ways. Only when the systems that extend our agency and act on our behalf are sourced with reliable records, can we make sense of them, take responsibility for them, and employ them with confidence.</p><p>All of this points to the viability of developing a software ecosystem in which open data standards and data in the form of private and shared logs of first-class user-instrumented events replaces application-internal data silos, potentially providing a great deal of data to peers for for training agents that automate processes. Such expression and control can be directly situated in and integrated into agents’ lived situations. By specifying agent-centered control as data, there is an opportunity for optimization of control, learning from past outcomes. While the term <strong>situated programming</strong> does not yet have much currency¹, it is a very fitting one to describe this trend of users gaining more of the power of programming and programmers gaining more situational context.</p><p>A final major question is the concern of the next chapter: what approach to information would allow software to helpfully participate meaningfully in the real situations that human participants control together. In other words how do we build bicycles for the mind? This needs to be a central concern of situated programming, I will argue. Scaling social organization requires expanding our capacities to share common ground, and common ground is established through the use of tools and and practices for coordinating attending to immediate concerns together.</p><h3>THE ART OF BICYCLE MAINTENANCE</h3><p>The proverbial bicycle for the mind can be defined more precisely as the minimal instrumentation of joint attentional skills required to establish machine-followable common ground. What I propose as way of constructing bicycles for the mind within the broader context of situated programming, is an approach to specifying the coordination of joint attention, which I call <strong>narrative process scaffolding</strong> (NPS). I have begun work on a protocol for agent-centered coordination on joint attentional scaffolding I call <strong>senters</strong> (a portmanteau of “sense-making” and “center”) as a way of implementing NPS. Here I will describe NPS and senters at a high level.</p><p>To review what was developed in the discussion above, there are two levels of meaning when it comes to agent-centered information. There is basic, first-order causative involvement (direct perception/action), and there is second-order convention-based (gesture/scaffolding) re-organization of first-order activity by controlling the selection, composition, and direction of activities of attending to the world. The approach of NPS is to <strong>instrument</strong> activities of direct world-involvement, and to <strong>scaffold</strong> convention-based processes of joint attentional orientation.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*IEq6bW25ZLomSFhzDOrbiw.png" /></figure><p>What are called <strong>instruments</strong> are typically simple functions that affect the world, including visual and haptic interactions for example. Users instrument their active involvement in environments, in an event-driven dataflow¹ style: instruments produce events when they are applied, on a feed specific to the agent’s use of the instrument. Such events are in turn referred to on feeds for what are called <strong>centers</strong> when the instrumented events are indicated by means of <strong>gestures</strong> as bearing on either attending to, supporting, or realizing the processes that a center scaffolds. The six gesture types cover empirically documented classes of engage/refrain (attending), follow/avoid² (supporting), and continue/exit (realizing). Gestures signal the relevance of events to attending toward ends. For example given an instrumented haptic event, gesturing “engage” for a center for moving an object means that event indicates an option to attend to that process. Centers accrete information that <strong>proxy agents</strong> follow to extend and support users’ actions.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*DH5JZkP205iU_LNd9DLf2A.png" /></figure><p>This approach allows for moving beyond an application model to a more composable and adaptable approach. Users leverage effective crowd-sourced software behaviors custom-tailored to fit real situations. The phrase “turning the application inside out”³ is a good description of the fact that centers (which are defined and shaped through use) replace the former role of applications, in capturing what functions are are available, how they compose, and where they get their users.</p><h3>CENTERS: ROADMAPS FOR BICYCLES FOR THE MIND</h3><p>The scientific hypothesis underwriting the technical proposal for narrative process scaffolding is that something we’ve termed a “center” describes and specifies the hard data of convention-based guidance of attention, or “scaffolding”. This concept is actually quite simple, but is subtle enough to be missed without a careful presentation, so this final chapter makes that presentation its objective. The title of the chapter asserts that centers are “roadmaps for bicycles of the mind”. Centers are the maps, not the territories that are mapped; the territory in this metaphor is the continuous flow of situated and embodied direct experience of participating in what happens.</p><p>In the following examples I use the descriptor “top-down” for any development/use where a center is engaged as scaffolding to direct a series of events, and “bottom-up” for development/use where occurrent events induce an appropriate center to scaffold contending with and making sense of the events. Also, a word on “development” versus “use”: use of centers is technically not distinct from enacting entirely new centers, as any use of a center revisits its story of development, and the record of its process can potentially scaffold a future episode of use by someone it is shared with. There is a practical difference of degree rather than kind in leveraging (at one extreme) a well-trodden center that one’s present experiences add nothing new to, and (at the other extreme) a center that has yet to be formed with respect to histories of events.</p><p><strong>Example 1. top-down development and use of centers.</strong> Alice, Bob, and some of their friends go on a hike and pick a new kind of berries. None of them are experts at picking berries. Alice and Bob individually draw from some centers that scaffold foraging generally, including picking berries of known varieties. Together the group collaborates on the development of centers to scaffold processes of acquiring this new kind of berry, benefiting from and refining one another’s findings as they progress, sharing with one another what works well.</p><p>You have been hiking in the woods for days with no rations, but you have a solar charging unit for your wearables. You access the local woodland mesh network and find that many foragers have shared scaffolding on their techniques. Your reserves of agents dutifully apply these centers to direct your attention to finding berries, and other food. There are also some relatively recently sourced centers scaffolding the process of finding a new kind of berry you have never heard of, but you don’t follow it, because of how few participants have contributed, and also because these pseudonymously signed strangers (Alice, Bob, and their friends) have made it difficult to link their activities to much else.</p><p><strong>Example 2. bottom-up development and use of centers.</strong> You are sitting outside in front of your home reading a book when suddenly a car’s screeching brakes break your concentration. A bicyclist flies from impact with the vehicle’s windshield. After seeing to it that the bicyclist receives medical care, you and the other nearby witnesses recount what you witnessed, and develop a consistent account, pooling pieces of recorded evidence, acts of claim-making (witnesses’, the artificial driver agent’s, the bicyclist’s) to create a content-bearing executable narrative that explains, in a process of joint attentional replay what transpired. It is accepted and applied as the basis of a “smart contract” insurance payout.</p><p><strong>~ EPILOGUE ~</strong></p><p>As these examples show, centers are histories of instrumented events directed by joint attentional guidance. Skillfully engaging centers means being able to use them to scaffold novel circumstances they apply to, following them through in episodes of attending toward realized outcomes. Many new ecologies of bicycles of the embodied mind can be discovered as we share, develop, and use these maps to experiencing them together.</p><h3>ENDNOTES</h3><p><strong>preface:</strong> <strong>1.</strong> To the best of my knowledge, the first documented occurrence of the turn of phrase was in a 1980 presentation by Apple Computer co-founder Steve Jobs on using computers to amplify human abilities. In so doing he paid homage to a key theme of the cybernetic movement that gave birth to computers. The title of this manifesto combines this phrase with a reference to “Steps to an Ecology of Mind” by cyberneticist Gregory Bateson, which explored the organization of individuals, societies, and ecosystems in terms feedback and adaptive self-regulation. <strong>2.</strong> In addition to bicycles standing for a close fit between humans and technological extension, bicycles have other fitting associations. Bicycles influence urban planning toward human-centered designs. They have historically stood for liberation and empowerment, playing a significant role in feminism. Susan B. Anthony wrote of the bicycle: “it has done more to emancipate women than anything else in the world. I stand and rejoice every time I see a woman ride by on a wheel.” <strong>3.</strong> Spoken language, by any account a crowning achievement of our species (innateness debates aside), is supported by image schematic and metaphorical devices for coordinating construal of abstract domains (Lakoff 1993, Talmy 2000, Goldberg 2003). The use of these devices is so entrenched in cultural practice that our use presents a task of unraveling mysteries for years to come in cognitive linguistics research.</p><p><strong>agents:</strong> <strong>1.</strong> Roughly 150 years ago a revolution in evolutionary theory occurred, most notably punctuated by Darwin’s “On the Origin of Species”. In that setting, the “pragmatic turn” in philosophy and science began with Charles Sanders Peirce’s project of reconstituting knowledge, meaning, and phenomenology in terms of situated agents calibrating meanings externally in networks of causative consequence. The pragmatic thread can be traced through the work of Uexkull, Vygotsky, later Wittgenstein, Heidegger, Merleau-Ponty and many others. Crucially it led to the 1st and 2nd cybernetics movements, the latter of which in conjunction with James and Eleanor Gibson’s pioneering work in ecological psychology has developed into the “e-turn” or “4E” (embodied, ecological, enactive, and extensive) cognitive sciences. The underlying theoretical basis of semiotics (Peircean, not Sassurian), cybernetics, and radical embodied cognitive science is that meaning happens in the world, relative to the predicaments of agents, and as such must be understood in terms of processes of agents’ involvement in the world over time. <strong>2.</strong> The assertion that agents are necessarily and sufficiently defined on three dimensions of situated bearing is mainly based on the work of Michael Tomasello, Adele Goldberg, and other related work in their usage-based program of communication research. This research has identified three ways in which humans learn (around nine months of age) to intervene in others’ processes of attending: sharing, following into, and directing processes of attending. These three skills of attention are directly related to three identified communicative motives of sharing, informing (supporting), and requesting (directing) process. These appear to be constitutive factors of the human semantic niche of specifying shared intention. In this manifesto I have taken the speculative liberty of extrapolating the historical themes of attention, work, and organization from these communicative skills or motives, as it seems to carry its explanatory weight. <strong>3.</strong> Turvey 1992 also supports this account of prospective awareness, supporting subacts, and afforded ends. <strong>4.</strong> Eleanor Gibson and Nancy Rader in “Attention: The Perceiver as Performer” (1979) explain that it would be misleading to regard attention as a special mechanism to be studied, and is more appropriately understood to be the activities of the perceiver-actor. Attending is simply the conduct of world-involving, end-directed activity. The word is however useful in pointing out that an agent must prioritize what to engage in, at the cost of engaging in some other activity. The word becomes more of a reference to a distinctive phenomenon in the discussion of joint attention, the phenomenon of agents’ states of attending being made targets of attention in their own right, to be influenced and organized. <strong>5.</strong> A case for maximizing potential next options as a default utility function for intrinsic motivation was first made by Klyubin, Polani, and Nehaniv 2005, which they term “empowerment”, quantifying it in information-theoretic terms as the channel capacity of an agent’s actuation channel terminating at the sensor.</p><p><strong>joint attention:</strong> <strong>1.</strong> Michael Tomasello’s research over several decades shows that directing the flow and focus of others’ attention in service of cooperative organization is the reason we direct the flow, focus, and organization our own world-involving activities. Basic ecological dynamics cannot be equivocated with the reflective reorganizational process of developing the what, how, and why of activity developed from from skills of joint attention. Some related work that builds this case for a duplex account of sapient intelligence are Hutto and Myin’s “Radicalizing Enactivism” and “Evolving Enactivism”, Chapman and Agre’s “Abstract Reasoning as Emergent from Concrete Activity”, Alva Noe’s “Strange Tools”, Nancy Salay’s “Representation: Problems and Solutions”, and Sabrina Golonka’s “Laws and conventions in language‐related behaviors”. Importantly, all of this work acknowledges that this duplex account is usage-based, at the level of embodied agents situated in the world together. Representational content is achieved by agents by means of the tools and practices they use to scaffold jointly attending. <strong>2.</strong> Ape gestures include “intention movements” (abbreviated actions that have effects within social contexts) and “attention getters” (actions that call attention, such as slapping the ground). Perhaps surprisingly, dogs (having had some time to adapt to our cognitive niche) have experimentally shown far better understanding of human pro-social intentions and gestures such as pointing than other animals (Kirchhofer, et al). <strong>3.</strong> The use of the term “narratively reorganized agency” refers to what is discussed in the first endnote of this section. Narrative process is a kind of feedback process concerned specifically with the selection, composition, and direction of what is attended to. As such, it is a second-order mode of end-directed feedback that is convention-based rather than law-based in its relation to ecological information. <strong>4.</strong> As Tracy Harms puts it, “mind, as cognitive scientists and philosophers focused on human flexibility are most interested, is characterized by an expanse of reorganizational fluidity.” Hutto et al (2014), on the limits of this extent write “limits can be revealed by empirical experiment. Even if one were able to see for miles one might not be able to see for leagues.”</p><p><strong>situated programming:</strong> <strong>1.</strong> William Clancey explored “situated action” in the context of computing in his influential 1997 book “Situated Cognition: On Human Knowledge and Computer Representation”. In 2002, Mills and Scholtz published “Situated Computing: The Next Frontier for HCI Research”, exploring the reorientation to a data-first, agent-centered approach to computing brought about by mobile and wearable devices. They write, “a future where people will interact with information through a continuously varying array of devices that combine to form ad hoc portals suitable to particular situations. In such a future, people and information will be emancipated.” The term “situated programming” was used by Rich Hickey in his keynote lecture at Clojure Conj 2017, emphasizing effective programs, the importance of developing meaning at the systems level rather than fixating on machine-level program abstraction, and the agent-centered design of systems.</p><p><strong>bicycle maintenance:</strong> <strong>1.</strong> Dominic Tarr’s secure scuttlebutt protocol (ssb) is an example of the sort of approach I have in mind. It applies the kappa architecture or event-driven dataflow programming at the user level, and for that reason it is a good fit for NPS. Further implementation-level concerns are explored in the senters RFC. <strong>2.</strong> “Avoid” might also be substituted with “voice” in reference to the notion of “voice and exit”. In the context of the support predicament of a narrative process, some other process can either be followed as supporting, or voiced as not supporting. In the realization predicament of a narrative process, “continue” and “exit” apply. <strong>3.</strong> To paraphrase Martin Kleppman’s (2014) “turning the database inside out” in reference to event sourcing. More broadly speaking this approach is in the spirit of Licklider’s (1960) vision of designing systems that “cooperate in making decisions and controlling complex situations without inflexible dependence on predetermined programs”.</p><p><strong>Acknowledgements:</strong> Dedicated to Bristy my partner in life and in joint attention. <em>Special thanks to Thomas Levy and Tracy Harms who in particular have spent countless hours both critically and constructively influencing the ideas presented here, and to Michael Travers, Reid McKensie, Brian Marick, Panicks Godeck, Simon Belak, Lexi Linnell, Zach Tellman, Kovas Boguta, Tianxiang Xiong, Brian Rice, Alyssa Kwan, Radford Smith, Max Kreminski, Damien Donnelly, Timothy Washington, David Nicol, Mia Lehrer, Alvaro Videla, Alex Hirner, Stephen Cagle, Eric Normand, Doug Haber, Rex Parker, Tim Howe and many others!</em></p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=95fa649ef422" width="1" height="1" alt="">]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[“Temporal Higher-Order Contracts” with Clojure.spec]]></title>
            <link>https://medium.com/@rplevy/temporal-higher-order-contracts-with-clojure-spec-e92e795665d?source=rss-3c03111694de------2</link>
            <guid isPermaLink="false">https://medium.com/p/e92e795665d</guid>
            <category><![CDATA[clojure]]></category>
            <category><![CDATA[contracts]]></category>
            <dc:creator><![CDATA[Robert P. Levy]]></dc:creator>
            <pubDate>Sat, 11 Jun 2016 04:11:38 GMT</pubDate>
            <atom:updated>2016-06-11T05:15:39.467Z</atom:updated>
            <content:encoded><![CDATA[<p>A sizable group of engineers including myself over at Funding Circle (a Clojure and Clojurescript shop, we’re hiring!) have been gathering every couple of weeks over lunch to discuss a paper that one of us has decided to present to the group. Yesterday, the paper we discussed was <a href="https://users.soe.ucsc.edu/~cormac/papers/icfp11.pdf">Temporal Higher-Order Contracts</a> presented by <a href="https://twitter.com/drb226">Dan Burton</a> who previously worked with one of the authors of the paper. In the discussion that followed, the question of the relation between Clojure.spec and contract systems came up. Clojure.spec in itself is arguably a complete contract system. But what additional tooling would be required to instrument contracts over temporal features of software systems? I decided to take a very shallow stab at addressing this.</p><p>A higher-order contract, as the term appears to be used in the Racket community, is a contract not just on a module or function, but on the contextual application of that module or function by some other. A temporal contract, higher-order or otherwise, is a contract that checks the behavior of a function or module over time, assuming a stateful scenario. The paper I mentioned approaches the implementation of temporal higher-order contracts in what seems like a very theoretically comprehensive and general way. Here I’d just like to illustrate limited functionality of this kind, perhaps to spark further discussion about what a useful generalization might look like.</p><p>To start using clojure.spec presently requires using the latest alpha release of Clojure, found at the top of <a href="https://github.com/clojure/clojure/releases">https://github.com/clojure/clojure/releases</a>. If you are reading this after the release of Clojure 1.9, just use the current stable release.</p><p>In the REPL, require clojure.spec:</p><pre>(require ‘[clojure.spec :as s])</pre><p>In the paper, the authors created a monitor to instrument all operations that are executed. Clojure.spec includes fdef, instrument, and unstrument for this purpose, but since we’ll want to instrument a function that might not be assigned to a Var, we’ll just write our own code to make some noise when something goes wrong that lets us know exactly what is responsible for the problem (yes, more informative Clojure stack traces!)</p><pre>(defn enforce [spec data-value]<br>  (let [r (s/conform spec data-value)]<br>    (if (= r :clojure.spec/invalid)<br>      (throw (ex-info (str r) (s/explain-data spec data-value)))<br>      data-value)))</pre><pre>(defn call-with-temporal-contract [f g contract-on-g]<br>  (let [function-calls (atom [])]<br>    (let [g&#39; (fn [&amp; [arg]]<br>               (swap! function-calls conj arg)<br>               (enforce contract-on-g @function-calls)<br>               (g arg))]<br>      (f g&#39;))))</pre><p>The enforce function is just a way of breaking program flow in an informative way when some behavior doesn’t conform to a contract. It does this by passing the output of clojure.spec/explain-data in the exception that it raises so that it can be found in the stack trace.</p><p>The call-with-temporal-contract function gives us, in a very non-general way, the ability to place a contract on the behavior of the function g over time as limited to its application in the scope of this one call to function f that uses g. For simplicity, g is assumed to take at most one argument, but it would be trivial to remove that simplifying assumption. The monitoring apparatus we’ve imposed on the higher-order function f creates an atom to collect call info from the instrumented version of the function g. This allows the instrumented function to incrementally, on each successive call of function g, apply a contract assessing the recorded history of times the function was applied in the scope of this call to f.</p><p>Let’s look at some examples. Here g is the print function “prn”, and we want to make sure that f doesn’t call it more than twice. But in fact f calls g three times.</p><pre>;; valid example<br>(call-with-temporal-contract<br>  (fn [g] (g “foo”) (g “bar”))<br>  prn<br>  #(&lt; (count %) 3))</pre><pre>;; invalid example<br>(call-with-temporal-contract<br>  (fn [g] (g “foo”) (g “bar”) (g “baz”))<br>  prn<br>  #(&lt; (count %) 3))</pre><p>In the stack trace for the invalid example you will see</p><pre>clojure.lang.ExceptionInfo: :clojure.spec/invalid {:clojure.spec/problems {[] {:pred :clojure.spec/unknown, :val [&quot;foo&quot; &quot;bar&quot; &quot;baz&quot;], :via [], :in []}}}</pre><p>Ideally we would rather see something about the fact we expected it to run less than 3 times, but it ran only 2. What this shows us is only the sequence of arguments to g that produced an invalid result. An actual implementation of temporal contracts would hopefully do better.</p><p>Now lets’s try a contract that specifies a sequence of states denoted by the allowed argument types that f can pass to g.</p><pre>;; valid example<br>(call-with-temporal-contract<br>  (fn [g] (g “foo”) (g :baz) (g 1))<br>  prn<br>  (s/cat :call1 string?<br>         :call2 (s/? keyword?)<br>         :call3 (s/? integer?)))</pre><pre>;; invalid example<br>(call-with-temporal-contract<br>  (fn [g] (g “foo”) (g :baz) (g :bar))<br>  prn<br>  (s/cat :call1 string?<br>         :call2 (s/? keyword?)<br>         :call3 (s/? integer?)))</pre><p>In the two examples above we apply the same rules evaluating the types of the successive arguments that our f function passes to its g function (the prn function) but in the latter invalid case, the third call breaks from the expected pattern. In the stack trace we find:</p><pre>clojure.lang.ExceptionInfo: :clojure.spec/invalid {:clojure.spec/problems {[:call3] {:pred integer?, :val :bar, :via [], :in [2]}}}</pre><p>In this case, we get more useful information, mostly because the features of spec (or my present level of proficiency with spec, this being one of my first times playing with it) fit the problem more closely. We can see that it was specifically the third call that failed, as it explicitly reports a problem with :call3. And what is more, we can see that it failed because it was supposed to be an integer, but the value was :bar (a keyword.)</p><p>From this comparatively crude approximation of something resembling temporal higher-order contracts, we can see that even with very little effort put into it, Clojure.spec can be applied to novel type of contract, adding expressiveness and better stack traces. With more thoughtful design and implementation, there is an opportunity to co-opt some powerful ideas on contracts found largely in Racket-based academic circles, and bring them into production-grade Clojure development.</p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=e92e795665d" width="1" height="1" alt="">]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[Cybernetics Versus Informatics: Understanding the Type Wars]]></title>
            <link>https://medium.com/@rplevy/cybernetics-versus-informatics-understanding-the-type-wars-99878fe956d6?source=rss-3c03111694de------2</link>
            <guid isPermaLink="false">https://medium.com/p/99878fe956d6</guid>
            <category><![CDATA[machine-learning]]></category>
            <category><![CDATA[artificial-intelligence]]></category>
            <dc:creator><![CDATA[Robert P. Levy]]></dc:creator>
            <pubDate>Tue, 31 May 2016 06:25:34 GMT</pubDate>
            <atom:updated>2016-05-31T06:55:32.920Z</atom:updated>
            <content:encoded><![CDATA[<h4>Two competing theories of the fundamental nature of system organization underly this contentious debate in computing</h4><p>The early history of computers was both a story of mathematical and electrical engineering achievements, and because of this, early there emerged two distinct but interacting cultures of computing. These two traditions were reflected in the first high-level programming languages Lisp and Fortran, Lisp being a product of mathematical interests, and Fortran the interests of engineers seeking to efficiently control machines with greater ease.</p><p>This article is not about the differences between mathematical purists and technical pragmatists (as there has seldom been a clear division, with each borrowing from the other.) Our focus here is on the theoretical orientations of informatics and cybernetics that have found affinity with the mathematical and technical cultures respectively. Note that this is not to imply that cybernetic approaches don’t make use of math, but rather to suggest that the horse pulls the cart and not vice versa. These two approaches remain at odds with one another, are at the heart of what we have come to regard as the “<a href="http://blog.cleancoder.com/uncle-bob/2016/05/01/TypeWars.html">The Type Wars</a>”. While the terminology as it has been used in the field is somewhat inconsistent, it is more or less clear to use the term <em>informatics</em> to refer to the goal of<strong> systems that consume data, manipulate it in terms of its semantic meaning, and produce meaningful results</strong>, and <em>cybernetics</em> to refer to the goal of <strong>systems that satisfy requirements meaningfully and adaptively in novel and uncertain environments</strong>.</p><p>In addition to bringing some clarity to this often obfuscated conflict underlying the type wars, I will try to present a convincing argument that while category theory may help our craft in limited but important ways, the greater aspirations of informatics are not likely to ever be delivered on, in computing or in cognitive science. Cybernetic approaches in contrast show a great deal of promise, and the current wave of progress in model-free AI and machine learning (notably <em>not</em> in computationalist methods) underscores this.</p><h4><strong>The Informatics Tradition in Computing and Psychology</strong></h4><p>Lisp being an elegant and expressively powerfully language found a niche in early forays into artificial intelligence, and became associated with the idea of symbolic computation, or computing with mathematical expressions. Symbolic computation was the province of computationalism, the theory that the brain is a computer, and that thought and computation are one and the same. Relatedly, and often in connection with AI and computational cognitive modeling efforts, cognitive psychologists began to adopt a neo-Cartesian philosophy of “information processing” borrowed from computing, seeking mechanisms by which raw sensory data is converted into a representation of the world, as it is processed and passed from brain module to brain module.</p><p>While the failure of the information processing model to significantly advance our understanding of cognition (more on this below) has been more gradual and protracted, the application of these ideas in computing famously collapsed under the weight of promises it could not deliver on. Unfortunately for the broader adoption and use of Lisp as a language, the close association with this naively ambitious and ultimately failed experiment in industry and academia meant that the AI Winter was also a cold time period for Lisp.</p><p>Even as symbolic AI collapsed and major problems were found with the theory of information processing, another thread of the mathematical tradition in computing that is so old as to pre-date the actual existence of computers, quietly continued to grow and improve: static type checking and formal program verification. Finding its roots in type theory and category theory, languages such as ML, Miranda, Haskell, and Coq leverage formal proof as tool for improving code quality.</p><p>While this might seem nothing more nor less than a modest and practical approach to safely developing software, today the proponents of the “programs are proofs” school aspire to epic ambitions rivaling those of the first wave of artificial intelligence. The highly influential <a href="http://math.ucr.edu/home/baez/rosetta.pdf">Rosetta Stone paper</a> for example asserts that the equivalence of propositions with types, physical systems, and topological manifolds, all encompassed by the object abstraction of category theory, points to the development of a “general science of systems and processes”. Far from a mere practical concern, category theorists have resurrected the information processing approach. This time we will successfully model the world with software, so they say. This time, they claim, we’ll successfully build software that captures meanings in representations, and by processing these representations we will extract and produce other useful meanings.</p><h4><strong>The Cybernetic Tradition in Systems Biology and Computing</strong></h4><p>Perhaps ironically, or perhaps due to the success of a small remnant Lisp community’s deliberate attempts to shake off old associations over the course of the 90s and 00s, Lisp came to be associated with a new narrative. As Lisp regained some popularity in the early 00s it came to be regarded as the alternative to an overly purist kind of functional programming preoccupied with types and proofs. Lisp in effect, switched sides to the other culture of computing — not in the sense of static vs. dynamic (it had always been dynamic) but in the sense of being positioned opposite the objectives of informatics, and playing up its strength as a dynamic language over its strengths in symbolic processing. After all, it had gained a reputation as a flexible language for things like evolutionary computation, as a diverse multi-paradigm set of building materials for domain-specific languages, and in general as a powerful tool enabling one to take on ambitious problems of all kinds.</p><p>By the end of the decade, Rich Hickey, an outspoken critic of the many of the claims made by static typing advocates, had succeeded in delivering Clojure and building a larger Lisp community than the world had ever before witnessed. Clojure also brought many good ideas into Lisp from other functional languages, such immutable data structures and laziness. In May of 2016, Rich Hickey announced <a href="https://clojure.org/about/spec">clojure.spec</a>, integrating a specification system into the core of the language, and took the opportunity to say:</p><blockquote>expressivity &gt; proof</blockquote><blockquote>There is no reason to limit our specifications to what we can prove, yet that is primarily what type systems do. There is so much more we want to communicate and verify about our systems. This goes beyond structural/representational types and tagging to predicates that e.g. narrow domains or detail relationships between inputs or between inputs and output. Additionally, the properties we care most about are often those of the runtime values, not some static notion. Thus <strong>spec</strong> is not a type system</blockquote><p>Doubtless some read this paragraph in clojure.spec’s stated rationale as nothing more than an opinionated jab at the programming language competition, but that would ignore what is actually going on here. This is a declaration of Clojure doubling down on it’s commitment to a philosophy of expressiveness and robustness through adaptiveness, in contrast with a philosophy of pre-conceived proof and self-contained meaning independent of runtime environments.</p><p>In the <a href="http://blog.cleancoder.com/uncle-bob/2016/05/01/TypeWars.html">“Type Wars”</a> article we referred to above, Bob makes a case that TDD (and presumably generative testing) is what dynamic languages have to offer in place of a type system. While the truth of this is evident to those who have experience with it, what might not be readily evident is that this is one of many examples of the cybernetic tradition in software development. TDD (and BDD, etc) is an outside-in approach that emphasizes the runtime behavior of the system under test. The purpose is to make systems more robust to changing conditions. Tests can be viewed as regulating the behavior of code, keeping functional requirements in place as the code that produces behaviors evolves over time.</p><p>To explore further, it’s important to note here that the history of this approach of adaptive control or environmental shaping of self-organizing systems is not merely a practical convenience, but one that that has been developed over many decades of scientific investigation. The early work on cybernetics as early as the 1940s focused on adaptive behavior in physical systems such as machines, and was the precursor to later work in machine learning, such as reinforcement learning. The aim of cybernetics was to find general principles of control applying to all systems. By the 1970s, cybernetics began to focus more on systems biology and mechanisms of control in living systems. In this period, Francisco Varela coined the term “autopoiesis” to describe the phenomenon of systems capable of maintaining and reproducing themselves. The study of autopoietic mechanisms led to the the development of enactivism, the theory that perceptual experience is the fully constituted by sensorimotor skills of probing and effecting environments, thus constructing modes of access and presence.</p><p>Earlier we mentioned the failure of cognitive science to conceive of a way for raw sensory data to be converted into a representation of the world, such that it could be processed meaningfully by a computer-like brain, enabling intelligent decisions. We drew a direct line of comparison to same problem not having been solved in computing, despite major efforts in GOFAI and now in category theory. The enactivist cybernetic framework in contrast does not suffer from these problems. Let’s consider for example a dog trying to access a treat on the other side of a barrier. First of all, the dog will skillfully and flexibly exploit affordances of the environment, so there is no need to represent a world that one is directly present in. Secondly, the dog perceives a thing that she has learned will produce a rewarding outcome, and experiments with available affordances, such as the climbable aspects of the barrier, in order to get to the treat. The informatician would overcomplicate this by trying to devise a representation of the world in the dog’s brain, and then process it to produce a transformation of that representation that would generate an action plan, or something along those lines. In the enactive frame of reference the only thing that needs to be in the dog’s brain is some way of remembering what skills work effectively with what direct contexts of action, where knowing the contexts of action is on equal footing with any other skill. The dog applies sensorimotor skills that have worked in similar circumstances, and no representation or manipulation of representations is ever needed.</p><p>Given the problems of the informatics approach that cybernetic approaches do not suffer from, it seems likely that we will continue to solve such problems by developing better adaptive meta-heuristic frameworks, and applying them to everyday computing needs. The dream of adaptive software agents that we train rather than program is not a pipe dream so much as a likely outcome of current progress in areas of research like reinforcement learning. Indeed, recent results by Google’s DeepMind and others show major advancements in unsupervised control from high-dimensional sensory inputs. I’m placing a long-term bet against category theory as a silver bullet, and in favor of soft-computing frameworks in which we compose semi-autonomous adaptive agents to solve problems.</p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=99878fe956d6" width="1" height="1" alt="">]]></content:encoded>
        </item>
    </channel>
</rss>