Artificial Intelligence Methodology

Foundation models such as GPT-4 are fine-tuned to avoid unsafe or otherwise problematic behavior, so that, for example, they refuse to comply with requests for help with committing crimes or with producing racist text. One approach to fine-tuning, called reinforcement learning from human feedback, learns from humans' expressed preferences over multiple outputs. Another approach is constitutional AI, in which the input from humans is a list of high-level principles. But how do we deal with potentially diverging input from humans? How (...) can we aggregate the input into consistent data about "collective" preferences or otherwise use it to make collective choices about model behavior? In this paper, we argue that the field of social choice is well positioned to address these questions, and we discuss ways forward for this agenda, drawing on discussions in a recent workshop on Social Choice for AI Ethics and Safety held in Berkeley, CA, USA in December 2023. (shrink)

The irrational exuberance associated with contemporary artificial intelligence (AI) reminds me of Charles Dickens: "it was the age of foolishness, it was the epoch of belief" (cf. Nature Editorial, 2016; to get a feel for the vanity fair that is AI, see Mitchell and Krakauer, 2023; Stilgoe, 2023). It is particularly distressing—feels like yet another rerun of Seinfeld, which is all about nothing (pun intended); we have seen it in the 60s and again in the 90s. AI might have had (...) an inauspicious beginning, which may very well be the original sin, and our current predicament is the attendant divine condemnation. Being unable to stand it, however divine it may be, without further ado, I get to the genesis (instead of getting lost in the here-and-now of this or that so-called breakthrough of AI). Here's my unvarnished critique! (shrink)

Mathematicians and software developers use the word "function" very differently, and yet, sometimes, things that are in practice implemented using the software developer's "function", are mathematically formalized using the mathematician's "function". This mismatch can lead to inaccurate formalisms. We consider a special case of this meta-problem. Various kinds of agents might, in actual practice, make use of private memory, reading and writing to a memory-bank invisible to the ambient environment. In some sense, we humans do this when we silently subvocalize (...) thoughts about the actions we are taking (at least when the environment we're in is too primitive to probe the contents of our brains). Mathematical function formalizations of agents often ignore this ability. We show that in a general agent-environment framework (of which reinforcement learning is a special case), in a technical sense, such private memories do not enable qualitatively different agent behavior. (shrink)

In this paper, we pursue three general aims: (I) We will define a notion of fundamental opacity and ask whether it can be found in High Energy Physics (HEP), given the involvement of machine learning (ML) and computer simulations (CS) therein. (II) We identify two kinds of non-fundamental, contingent opacity associated with CS and ML in HEP respectively, and ask whether, and if so how, they may be overcome. (III) We address the question of whether any kind of opacity, contingent (...) or fundamental, is unique to ML or CS, or whether they stand in continuity to kinds of opacity associated with other scientific research. (shrink)

An important module in the Belief-Desire-Intention architecture for artificial agents (which builds on Michael Bratman's work in the philosophy of action) focuses on the task of intention reconsideration. The theoretical task is to formulate principles governing when an agent ought to undo a prior committed intention and reopen deliberation. Extant proposals for such a principle, if sufficiently detailed, are either too task-specific or too computationally demanding. I propose that an agent ought to reconsider an intention whenever some incompatible prospect is (...) sufficiently valuable along some dimension that can be assessed at zero or near-zero computational cost. (shrink)

The present paper looks at one of the most thorough articles on the intelligence of GPT, research conducted by engineers at Microsoft. Although there is a great deal of value in their work, I will argue that, for familiar philosophical reasons, their methodology, ‘Black-box Interpretability’ is wrongheaded. But there is a better way. There is an exciting and emerging discipline of ‘Inner Interpretability’ (also sometimes called ‘White-box Interpretability’) that aims to uncover the internal activations and weights of models in order (...) to understand what they represent and the algorithms they implement. In my view, a crucial mistake in Black-box Interpretability is the failure to appreciate that how processes are carried out matters when it comes to intelligence and understanding. I can’t pretend to have a full story that provides both necessary and sufficient conditions for being intelligent, but I do think that Inner Interpretability dovetails nicely with plausible philosophical views of what intelligence requires. So the conclusion is modest, but the important point in my view is seeing how to get the research on the right track. Towards the end of the paper, I will show how some of the philosophical concepts can be used to further refine how Inner Interpretability is approached, so the paper helps draw out a profitable, future two-way exchange between Philosophers and Computer Scientists. (shrink)

For synergistic interactions between humans and artificial intelligence (AI) systems, AI outputs often need to be explainable to people. Explainable AI (XAI) systems are commonly tested in human user studies. However, whether XAI researchers consider potential cultural differences in human explanatory needs remains unexplored. We highlight psychological research that found significant differences in human explanations between many people from Western, commonly individualist countries and people from non-Western, often collectivist countries. We argue that XAI research currently overlooks these variations and that (...) many popular XAI designs implicitly and problematically assume that Western explanatory needs are shared cross-culturally. Additionally, we systematically reviewed over 200 XAI user studies and found that most studies did not consider relevant cultural variations, sampled only Western populations, but drew conclusions about human-XAI interactions more generally. We also analyzed over 30 literature reviews of XAI studies. Most reviews did not mention cultural differences in explanatory needs or flag overly broad cross-cultural extrapolations of XAI user study results. Combined, our analyses provide evidence of a cultural bias toward Western populations in XAI research, highlighting an important knowledge gap regarding how culturally diverse users may respond to widely used XAI systems that future work can and should address. (shrink)

Many ethical frameworks require artificial intelligence (AI) systems to be explainable. Explainable AI (XAI) models are frequently tested for their adequacy in user studies. Since different people may have different explanatory needs, it is important that participant samples in user studies are large enough to represent the target population to enable generalizations. However, it is unclear to what extent XAI researchers reflect on and justify their sample sizes or avoid broad generalizations across people. We analyzed XAI user studies (N = (...) 220) published between 2012 and 2022. Most studies did not offer rationales for their sample sizes. Moreover, most papers generalized their conclusions beyond their target population, and there was no evidence that broader conclusions in quantitative studies were correlated with larger samples. These methodological problems can impede evaluations of whether XAI systems implement the explainability called for in ethical frameworks. We outline principles for more inclusive XAI user studies. (shrink)

Inspired by recent progress in multi-agent Reinforcement Learning (RL), in this work we examine the collective intelligent behaviour of theoretical universal agents by introducing a weighted mixture operation. Given a weighted set of agents, their weighted mixture is a new agent whose expected total reward in any environment is the corresponding weighted average of the original agents' expected total rewards in that environment. Thus, if RL agent intelligence is quantified in terms of performance across environments, the weighted mixture's intelligence is (...) the weighted average of the original agents' intelligences. This operation enables various interesting new theorems that shed light on the geometry of RL agent intelligence, namely: results about symmetries, convex agent-sets, and local extrema. We also show that any RL agent intelligence measure based on average performance across environments, subject to certain weak technical conditions, is identical (up to a constant factor) to performance within a single environment dependent on said intelligence measure. (shrink)

Can an ugly experiment be a good experiment? Philosophers have identified many beautiful experiments and explored ways in which their beauty might be connected to their epistemic value. In contrast, the present chapter seeks out (and celebrates) ugly experiments. Among the ugliest are those being designed by AI algorithms. Interestingly, in the contexts where such experiments tend to be deployed, low aesthetic value correlates with high epistemic value. In other words, ugly experiments can be good. Given this, we should conclude (...) that beauty is not generally necessary or sufficient for epistemic value, and increasing beauty will not generally tend to increase epistemic value. (shrink)

This is a paper about the general theory of measuring or estimating social intelligence via benchmarks. Hernández-Orallo and Dowe described a problem with certain proposed intelligence measures. The problem suggests that those intelligence measures might not accurately capture social intelligence. We argue that Hernández-Orallo and Dowe's problem is even more general than how they stated it, applying to many subdomains of AGI, not just the one subdomain in which they stated it. We then propose a solution. In our solution, instead (...) of using test-cases within the given AGI subdomain to estimate an AI's intelligence, one would use test-cases in an extended subdomain where test-cases have the ability to simulate the AI being tested. Surprisingly, AIs only designed for the original subdomain can be tested with test-cases in the extended subdomain anyway. By extending the subdomain in this way, we might avoid Hernández-Orallo and Dowe's problem. (shrink)

We consider an extended notion of reinforcement learning in which the environment can simulate the agent and base its outputs on the agent's hypothetical behavior. Since good performance usually requires paying attention to whatever things the environment's outputs are based on, we argue that for an agent to achieve on-average good performance across many such extended environments, it is necessary for the agent to self-reflect. Thus weighted-average performance over the space of all suitably well-behaved extended environments could be considered a (...) way of measuring how self-reflective an agent is. We give examples of extended environments and introduce a simple transformation which experimentally seems to increase some standard RL agents' performance in a certain type of extended environment. (shrink)

We present a game mechanic called pseudo-visibility for games inhabited by non-player characters (NPCs) driven by reinforcement learning (RL). NPCs are incentivized to pretend they cannot see pseudo-visible players: the training environment simulates an NPC to determine how the NPC would act if the pseudo-visible player were invisible, and penalizes the NPC for acting differently. NPCs are thereby trained to selectively ignore pseudo-visible players, except when they judge that the reaction penalty is an acceptable tradeoff (e.g., a guard might accept (...) the penalty in order to protect a treasure because losing the treasure would hurt even more). We describe an RL agent transformation which allows RL agents that would not otherwise do so to perform some limited self-reflection to learn the training environments in question. (shrink)

Although machine intelligence is increasingly employed in healthcare, the realm of decision-making in medical ethics remains largely unexplored from a technical perspective. We propose an approach based on fuzzy cognitive maps (FCMs), which builds on Beauchamp and Childress’ prima-facie principles. The FCM’s weights are optimized using a genetic algorithm to provide recommendations regarding the initiation, continuation, or withdrawal of medical treatment. The resulting model approximates the answers provided by our team of medical ethicists fairly well and offers a high degree (...) of interpretability. Possible applications of such a system include informal guidance on medical ethics dilemmas as well as educational purposes. (shrink)

As more objections have been raised against grant peer-review for being costly and time-consuming, the legitimate question arises whether machine learning algorithms could help assess the epistemic efficiency of the proposed projects. As a case study, we investigated whether project efficiency in high energy physics can be algorithmically predicted based on the data from the proposal. To analyze the potential of algorithmic prediction in HEP, we conducted a study on data about the structure and outcomes of HEP experiments with the (...) goal of predicting their efficiency. In the first step, we assessed the project efficiency using Data Envelopment Analysis of 67 experiments conducted in the HEP laboratory Fermilab. In the second step, we employed predictive algorithms to detect which team structures maximize the epistemic performance of an expert group. For this purpose, we used the efficiency scores obtained by DEA and applied predictive algorithms – lasso and ridge linear regression, neural network, and gradient boosted trees – on them. The results of the predictive analyses show moderately high accuracy, indicating that they can be beneficial as one of the steps in grant review. Still, their applicability in practice should be approached with caution. Some of the limitations of the algorithmic approach are the unreliability of citation patterns, unobservable variables that influence scientific success, and the potential predictability of the model. (shrink)

Intelligence in current AI research is measured according to designer-assigned tasks that lack any relevance for an agent itself. As such, tasks and their evaluation reveal a lot more about our intelligence than the possible intelligence of agents that we design and evaluate. As a possible first step in remedying this, this article introduces the notion of “self-concern,” a property of a complex system that describes its tendency to bring about states that are compatible with its continued self-maintenance. Self-concern, as (...) argued, is the foundation of the kind of basic intelligence found across all biological systems, because it reflects any such system's existential task of continued viability. This article aims to cautiously progress a few steps closer to a better understanding of some necessary organisational conditions that are central to self-concern in biological systems. By emulating these conditions in embodied AI, perhaps something like genuine self-concern can be implemented in machines, bringing AI one step closer to its original goal of emulating human-like intelligence. (shrink)

In their paper 'Reward is enough', Silver et al conjecture that the creation of sufficiently good reinforcement learning (RL) agents is a path to artificial general intelligence (AGI). We consider one aspect of intelligence Silver et al did not consider in their paper, namely, that aspect of intelligence involved in designing RL agents. If that is within human reach, then it should also be within AGI's reach. This raises the question: is there an RL environment which incentivises RL agents to (...) design RL agents? (shrink)

Can an agent's intelligence level be negative? We extend the Legg-Hutter agent-environment framework to include punishments and argue for an affirmative answer to that question. We show that if the background encodings and Universal Turing Machine (UTM) admit certain Kolmogorov complexity symmetries, then the resulting Legg-Hutter intelligence measure is symmetric about the origin. In particular, this implies reward-ignoring agents have Legg-Hutter intelligence 0 according to such UTMs.

In 2011, Hibbard suggested an intelligence measure for agents who compete in an adversarial sequence prediction game. We argue that Hibbard’s idea should actually be considered as two separate ideas: first, that the intelligence of such agents can be measured based on the growth rates of the runtimes of the competitors that they defeat; and second, one specific (somewhat arbitrary) method for measuring said growth rates. Whereas Hibbard’s intelligence measure is based on the latter growth-rate-measuring method, we survey other methods (...) for measuring function growth rates, and exhibit the resulting Hibbard-like intelligence measures and taxonomies. Of particular interest, we obtain intelligence taxonomies based on Big-O and Big-Theta notation systems, which taxonomies are novel in that they challenge conventional notions of what an intelligence measure should look like. We discuss how intelligence measurement of sequence predictors can indirectly serve as intelligence measurement for agents with Artificial General Intelligence (AGIs). (shrink)

Artificial intelligence (AI) research enjoyed an initial period of enthusiasm in the 1970s and 80s. But this enthusiasm was tempered by a long interlude of frustration when genuinely useful AI applications failed to be forthcoming. Today, we are experiencing once again a period of enthusiasm, fired above all by the successes of the technology of deep neural networks or deep machine learning. In this paper we draw attention to what we take to be serious problems underlying current views of artificial (...) intelligence encouraged by these successes, especially in the domain of language processing. We then show an alternative approach to language-centric AI, in which we identify a role for philosophy. (shrink)

We propose that, for the purpose of studying theoretical properties of the knowledge of an agent with Artificial General Intelligence (that is, the knowledge of an AGI), a pragmatic way to define such an agent’s knowledge (restricted to the language of Epistemic Arithmetic, or EA) is as follows. We declare an AGI to know an EA-statement φ if and only if that AGI would include φ in the resulting enumeration if that AGI were commanded: “Enumerate all the EA-sentences which you (...) know.” This definition is non-circular because an AGI, being capable of practical English communication, is capable of understanding the everyday English word “know” independently of how any philosopher formally defines knowledge; we elaborate further on the non-circularity of this circular-looking definition. This elegantly solves the problem that different AGIs may have different internal knowledge definitions and yet we want to study knowledge of AGIs in general, without having to study different AGIs separately just because they have separate internal knowledge definitions. Finally, we suggest how this definition of AGI knowledge can be used as a bridge which could allow the AGI research community to import certain abstract results about mechanical knowing agents from mathematical logic. (shrink)

Can an AGI create a more intelligent AGI? Under idealized assumptions, for a certain theoretical type of intelligence, our answer is: “Not without outside help”. This is a paper on the mathematical structure of AGI populations when parent AGIs create child AGIs. We argue that such populations satisfy a certain biological law. Motivated by observations of sexual reproduction in seemingly-asexual species, the Knight-Darwin Law states that it is impossible for one organism to asexually produce another, which asexually produces another, and (...) so on forever: that any sequence of organisms (each one a child of the previous) must contain occasional multi-parent organisms, or must terminate. By proving that a certain measure (arguably an intelligence measure) decreases when an idealized parent AGI single-handedly creates a child AGI, we argue that a similar Law holds for AGIs. (shrink)

After generalizing the Archimedean property of real numbers in such a way as to make it adaptable to non-numeric structures, we demonstrate that the real numbers cannot be used to accurately measure non-Archimedean structures. We argue that, since an agent with Artificial General Intelligence (AGI) should have no problem engaging in tasks that inherently involve non-Archimedean rewards, and since traditional reinforcement learning rewards are real numbers, therefore traditional reinforcement learning probably will not lead to AGI. We indicate two possible ways (...) traditional reinforcement learning could be altered to remove this roadblock. (shrink)

This article offers comprehensive criticism of the Turing test and develops quality criteria for new artificial general intelligence (AGI) assessment tests. It is shown that the prerequisites A. Turing drew upon when reducing personality and human consciousness to “suitable branches of thought” re-flected the engineering level of his time. In fact, the Turing “imitation game” employed only symbolic communication and ignored the physical world. This paper suggests that by restricting thinking ability to symbolic systems alone Turing unknowingly constructed “the wall” (...) that excludes any possi-bility of transition from a complex observable phenomenon to an abstract image or concept. It is, therefore, sensible to factor in new requirements for AI (artificial intelligence) maturity assessment when approaching the Tu-ring test. Such AI must support all forms of communication with a human being, and it should be able to comprehend abstract images and specify con-cepts as well as participate in social practices. (shrink)

The goal of creating Artificial General Intelligence (AGI) – or in other words of creating Turing machines (modern computers) that can behave in a way that mimics human intelligence – has occupied AI researchers ever since the idea of AI was first proposed. One common theme in these discussions is the thesis that the ability of a machine to conduct convincing dialogues with human beings can serve as at least a sufficient criterion of AGI. We argue that this very ability (...) should be accepted also as a necessary condition of AGI, and we provide a description of the nature of human dialogue in particular and of human language in general against this background. We then argue that it is for mathematical reasons impossible to program a machine in such a way that it could master human dialogue behaviour in its full generality. This is (1) because there are no traditional explicitly designed mathematical models that could be used as a starting point for creating such programs; and (2) because even the sorts of automated models generated by using machine learning, which have been used successfully in areas such as machine translation, cannot be extended to cope with human dialogue. If this is so, then we can conclude that a Turing machine also cannot possess AGI, because it fails to fulfil a necessary condition thereof. At the same time, however, we acknowledge the potential of Turing machines to master dialogue behaviour in highly restricted contexts, where what is called “narrow” AI can still be of considerable utility. (shrink)

The term ‘intelligence’ as used in this paper refers to items of knowledge collected for the sake of assessing and maintaining national security. The intelligence community (IC) of the United States (US) is a community of organizations that collaborate in collecting and processing intelligence for the US. The IC relies on human-machine-based analytic strategies that 1) access and integrate vast amounts of information from disparate sources, 2) continuously process this information, so that, 3) a maximally comprehensive understanding of world actors (...) and their behaviors can be developed and updated. Herein we describe an approach to utilizing outcomes-based learning (OBL) to support these efforts that is based on an ontology of the cognitive processes performed by intelligence analysts. Of particular importance to the Cognitive Process Ontology is the class Representation that is Warranted. Such a representation is descriptive in nature and deserving of trust in its veridicality. The latter is because a Representation that is Warranted is always produced by a process that was vetted (or successfully designed) to reliably produce veridical representations. As such, Representations that are Warranted are what in other contexts we might refer to as ‘items of knowledge’. (shrink)

Alcuni anni fa, ho raggiunto il punto in cui di solito posso dire dal titolo di un libro, o almeno dai titoli dei capitoli, quali tipi di errori filosofici saranno fatti e con quale frequenza. Nel caso di opere nominalmente scientifiche queste possono essere in gran parte limitate a determinati capitoli che sono filosofici o cercanodi trarre conclusioni generali sul significato o sul significato a lungoterminedell'opera. Normalmente però le questioni scientifiche di fatto sono generosamente intrecciate con incomprodellami filosofici su ciò (...) che questi fatti significano. Le chiare distinzioni che Wittgenstein ha descritto circa 80 anni fa tra le questioni scientifiche e le loro descrizioni da parte di vari giochi linguistici sono raramente prese in considerazione, e quindi si è alternativamente stupiti dalla scienza e costerghi per la sua analisi incoerente. Così è con questo volume. Se si vuole creare una mente più o meno come la nostra, è necessario avere una struttura logica per la razionalità e la comprensione dei due sistemi di pensiero (teoria del doppio processo). Se si vuole filosofare su questo, bisogna capire la distinzione tra le questioni scientifiche di fatto e la questione filosofica di come funziona il linguaggio nel contesto in questione, e di come evitare le insidie del riduzionismo e dello scientismo, ma Kurzweil, come la maggior parte degli studenti di comportamento, è in gran parte all'oscuro. Egli è incantato da modelli, teorie e concetti, e dalla voglia di spiegare, mentre Wittgenstein ci ha mostrato che dobbiamo solo descrivere, e che le teorie, i concetti, ecc., sono solo modi di usare il linguaggio (giochi linguistici) che hanno valore solo nella misura in cui hanno un test chiaro (clear truthmakers, o come John Searle (il critico più famoso di AI) ama dire, chiare condizioni di soddisfazione (COS)). Ho cercato di dare un inizio su questo nei miei scritti recenti. Coloro che desiderano un quadro aggiornato completo per il comportamento umano dalla moderna vista a due systems possono consultare il mio libro 'La struttura logica della filosofia, psicologia, mente e il linguaggio in Ludwig Wittgenstein e John Searle' 2nd ed (2019). Coloro che sono interessati a più dei miei scritti possono vedere 'Talking Monkeys--Filosofia, Psicologia, Scienza, Religione e Politica su un Pianeta Condannato--Articoli e Recensioni 2006-2019 3rd ed (2019) e Suicidal Utopian Delusions in the 21st Century 5th ed (2019) . (shrink)

Alcuni anni fa, ho raggiunto il punto in cui di solito posso dire dal titolo di un libro, o almeno dai titoli dei capitoli, quali tipi di errori filosofici saranno fatti e con quale frequenza. Nel caso di opere nominalmente scientifiche queste possono essere in gran parte limitate a determinati capitoli che sono filosofici o cercanodi trarre conclusioni generali sul significato o sul significato a lungoterminedell'opera. Normalmente però le questioni scientifiche di fatto sono generosamente intrecciate con incomprodellami filosofici su ciò (...) che questi fatti significano. Le chiare distinzioni che Wittgenstein ha descritto circa 80 anni fa tra le questioni scientifiche e le loro descrizioni da parte di vari giochi linguistici sono raramente prese in considerazione, e quindi si è alternativamente stupiti dalla scienza e costerghi per la sua analisi incoerente. Così è con questo volume. Se si vuole creare una mente più o meno come la nostra, è necessario avere una struttura logica per la razionalità e la comprensione dei due sistemi di pensiero (teoria del doppio processo). Se si vuole filosofare su questo, bisogna capire la distinzione tra le questioni scientifiche di fatto e la questione filosofica di come funziona il linguaggio nel contesto in questione, e di come evitare le insidie del riduzionismo e dello scientismo, ma Kurzweil, come la maggior parte degli studenti di comportamento, è in gran parte all'oscuro. Egli è incantato da modelli, teorie e concetti, e dalla voglia di spiegare, mentre Wittgenstein ci ha mostrato che dobbiamo solo descrivere, e che le teorie, i concetti, ecc., sono solo modi di usare il linguaggio (giochi linguistici) che hanno valore solo nella misura in cui hanno un test chiaro (clear truthmakers, o come John Searle (il critico più famoso di AI) ama dire, chiare condizioni di soddisfazione (COS)). Ho cercato di dare un inizio su questo nei miei scritti recenti. Coloro che desiderano un quadro aggiornato completo per il comportamento umano dalla moderna vista a due systems possono consultare il mio libro 'La struttura logica della filosofia, psicologia, mente e il linguaggio in Ludwig Wittgenstein e John Searle' 2nd ed (2019). Coloro che sono interessati a più dei miei scritti possono vedere 'Talking Monkeys--Filosofia, Psicologia, Scienza, Religione e Politica su un Pianeta Condannato--Articoli e Recensioni 2006-2019 3rd ed (2019) e Suicidal Utopian Delusions in the 21st Century 5th ed (2019) . (shrink)

पहली बात हमें ध्यान में रखना चाहिए कि जब यह कहना है कि चीन यह कहता है या चीन ऐसा करता है, तो हम चीनी लोगों की बात नहीं कर रहे हैं, लेकिन उन सोशियोपैथों की जो सीसीपी (चीनी कम्युनिस्ट पार्टी, अर्थात सात सेनेले सोसोपैथिक सीरियल किलर (एसएसएसएसके) का नियंत्रण करते हैं। सीपी या पोलितब्यूरो के 25 सदस्यों की टंडिंग समिति। मैं हाल ही में कुछ ठेठ वामपंथी नकली समाचार कार्यक्रमों को देखा (सुंदर बहुत ही तरह एक ही तरह से (...) मीडिया में पा सकते हैं, यानी, लगभग सब कुछ अब - यानी, याहू, सीएनएन, न्यूयॉर्क टाइम्स, आदि) यूट्यूब पर, एक VICE द्वारा जो उल्लेख किया है कि 1000 अर्थशास्त्री (और 15 नोबेल पुरस्कार विजेताओं) ट्रम्प को एक पत्र भेजा जिसमें कहा गया है कि व्यापार युद्ध एक गलती थी, और एक अन्य जो एक अकादमिक अर्थशास्त्री का साक्षात्कार करता था जिसने कहा कि ट्रम्प का कदम 3 विश्व युद्ध शुरू करने के लिए एक उत्तेजना थी। वे वैश्विक व्यापार के विघटन के बारे में सही हैं, लेकिन बड़ी तस्वीर है, जो है कि सात Sociopaths कुल दुनिया प्रभुत्व है की कोई समझ नहीं है, हर जगह स्वतंत्रता के उन्मूलन के साथ, अपने लक्ष्य के रूप में, और है कि वहाँ केवल दो तरीके उन्हें रोकने के लिए कर रहे हैं-एक कुल व्यापार प्रतिबंध है कि चीनी अर्थव्यवस्था devastates और उनकी सेना की ओर जाता है सीसीपी बाहर बल और चुनाव, या WW3, जो सीमित किया जा सकता है (शायद कुछ परमाणु के साथ पारंपरिक हथियार) या कुल (एक ही बार में सभी परमाणु). दिन के रूप में स्पष्ट है, लेकिन इन सभी "शानदार" शिक्षाविदों यह नहीं देख सकते हैं. यदि Socipaths अब नहीं हटा रहे हैं, के रूप में छोटे रूप में 15 साल में यह बहुत देर हो जाएगी, और अपने वंशजों को धीरे धीरे लेकिन inexorably चीनी के रूप में एक ही भाग्य के अधीन हो जाएगा -अपहरण, यातना और किसी भी असहमति के हत्या के साथ कुल निगरानी. बेशक, सीसीपी WW3 बहुत पहले शुरू कर दिया (आप तिब्बत या कोरिया के अपने आक्रमण की शुरुआत के रूप में देख सकता है) और यह हर संभव तरीके से पीछा कर रहा है, गोलियों और बम के अलावा, और वे जल्द ही आ जाएगा. सीसीपी ने कोरिया में संयुक्त राज्य अमेरिका से लड़ाई की, तिब्बत पर हमला किया और उसकी हत्या कर दी और रूस और भारत के साथ सीमा पर झड़पें लड़ीं। यह दुनिया भर में सभी औद्योगिक और सैन्य डेटाबेस के खिलाफ बड़े पैमाने पर हैकिंग आपरेशन आयोजित करता है और लगभग सभी वर्तमान अमेरिका और यूरोपीय सैन्य और अंतरिक्ष प्रणालियों पर वर्गीकृत डेटा चोरी हो गया है, उनकी कमजोरियों का विश्लेषण किया और भीतर सुधार संस्करण क्षेत्र कुछ साल. हजारों की दसियों, और शायद हजारों की सैकड़ों, सीसीपी कर्मचारियों की हैकिंग किया गया है सैन्य, औद्योगिक, वित्तीय और सामाजिक मीडिया डेटाबेस दुनिया भर में नेट के शुरुआती दिनों के बाद से और वहाँ अकेले संयुक्त राज्य अमेरिका में ज्ञात हाल ही में हैक्स के सैकड़ों रहे हैं. के रूप में प्रमुख संस्थानों और सेना उनके फायरवॉल कठोर है, SSSSK छोटे संस्थानों के लिए चले गए हैं और रक्षा उपठेकेदारों और हमारे सहयोगी दलों, जो आसान लक्ष्य हैं. हालांकि यह लाखों लोगों के सैकड़ों की कुचल गरीबी और अपने लोगों के सीमांत अस्तित्व की अनदेखी, यह एक बड़े पैमाने पर सैन्य और अंतरिक्ष उपस्थिति है, जो हर साल बड़ा बढ़ता है, और जिसका अस्तित्व के लिए ही कारण स्वतंत्रता को खत्म करने के लिए युद्ध लड़ रहा है बनाया गया है हर जगह. संसाधनों के 3rd दुनिया अलग करना के अलावा, बहु खरब डॉलर बेल्ट और सड़क पहल का एक प्रमुख जोर दुनिया भर में सैन्य ठिकानों का निर्माण कर रहा है. वे एक बड़े पैमाने पर उच्च तकनीक हथियारों की दौड़ है कि सोवियत संघ के साथ शीत युद्ध एक पिकनिक की तरह लग रही है में मुक्त दुनिया मजबूर कर रहे हैं. रूसके लोग बेवकूफ नहीं हैं, और Socipaths के साथ दोस्ती का नाटक करने के बावजूद, वे निश्चित रूप से समझते हैं कि सीसीपी उन्हें जिंदा खाने जा रहा है, कि उनकी एकमात्र उम्मीद खुद को पश्चिम के साथ सहयोगी है, और ट्रम्प पुतिन से दोस्ती करने में पैसे पर सही है। बेशक, Neomarxist तीसरी दुनिया Supremacists (यानी, डेमोक्रेटिक पार्टी) की संभावना में संयुक्त राज्य अमेरिका के कुल नियंत्रण ले जाएगा 2020 और कुछ भी नहीं सीसीपी की पसंद के लिए और अधिक हो सकता है. स्नोडेन (एक और सुराग बीस कुछ) SSSSK किसी भी अन्य एक व्यक्ति की तुलना में अधिक मदद की, WW2 के बाद से सभी अमेरिकी राष्ट्रपतियों के संभावित अपवाद के साथ, जो तुष्टीकरण की आत्महत्या की नीति का पीछा किया है. संयुक्त राज्य अमेरिका के पास सभी संचारों की निगरानी करने और हर किसी पर एक डोजियर संकलित करने के अलावा कोई विकल्प नहीं है, क्योंकि यह न केवल अपराधियों और आतंकवादियों को नियंत्रित करने के लिए आवश्यक है, बल्कि एसएसएसएसके का मुकाबला करने के लिए, जो तेजी से एक ही काम कर रहे हैं, पूरी तरह से स्वतंत्रता को हटाने के इरादे से। हालांकि SSSSK, और दुनिया की सेना के बाकी, उन्नत हार्डवेयर पर भारी रकम खर्च कर रहे हैं, यह अत्यधिक संभावना है कि WW3 (या छोटे सगाई यह करने के लिए अग्रणी) सॉफ्टवेयर प्रभुत्व होगा. यह सवाल से बाहर नहीं है कि SSSSK, शायद अधिक हैकर्स के साथ (कोडर्स) उनके लिए काम कर रहे तो संयुक्त दुनिया के सभी बाकी, कम से कम शारीरिक संघर्ष के साथ भविष्य के युद्ध जीत जाएगा, बस नेट के माध्यम से अपने दुश्मनों को paralyzing द्वारा. कोई उपग्रहों, कोई फोन, कोई संचार, कोई वित्तीय लेनदेन, कोई पावर ग्रिड, कोई इंटरनेट, कोई उन्नत हथियार, कोई वाहन, गाड़ियों, जहाजों या विमानों. कुछ सवाल हो सकता है कि सीसीपी (और निश्चित रूप से पुलिस, सेना और 610 कार्यालय के शीर्ष स्तरों) वास्तव में मानसिक रूप से aberant हैं, तो यहाँ sociopaths के आम विशेषताओं में से कुछ हैं(पूर्व में कहा जाता है psychopaths) कि आप नेट पर पा सकते हैं. बेशक, इनमें से कुछ कई autistics और alexithymics द्वारा साझा कर रहे हैं, और socipaths केवल डिग्री में "सामान्य" लोगों से अलग हैं. (shrink)

Nel 'Godel's Way' tre eminenti scienziati discutono questioni come l'indecidibilità, l'incompletezza, la casualità, la computabilità e la paracoerenza. Affronto questi problemi dal punto di vista di Wittgensteinian che ci sono due questioni fondamentali che hanno soluzioni completamente diverse. Ci sono le questioni scientifiche o empiriche, che sono fatti sul mondo che devono essere studiati in modo osservante e filosofico su come il linguaggio può essere usato in modo intelligibilmente (che include alcune domande in matematica e logica), che devono essere decise (...) cercando un modo in cui usiamoeffettivamente le parole in particolari contesti. Quando arriviamo a sapere su quale gioco di lingua stiamo giocando, questi argomenti sono visti come domande scientifiche e matematiche ordinarie come tutti gli altri. Le intuizioni di Wittgenstein sono state raramente eguagliate e mai superate e sono pertinenti oggi come lo erano 80 anni fa, quando ha dettato i Libri Blu e Marrone. Nonostante le sue mancanze - in realtà una serie di note piuttosto che un libro finito - questa è una fonte unica del lavoro di questi tre famosi studiosi che hanno lavorato ai bordi sanguinanti della fisica, della matematica e della filosofia per oltre mezzo secolo. Da Costa e Doria sono citati da Wolpert (vedi sotto o i miei articoli su Wolpert e la mia recensione di 'I limiti esterni della ragione' di Yanofsky) dal momento che hanno scritto sul calcolo universale, e tra i suoi numerosi successi, Da Costa è un pioniere della paracoerenza. Coloro che desiderano un quadro aggiornato completo per il comportamento umano dalla moderna vista a due systems possono consultare il mio libro 'La struttura logica dellafilosofia, psicologia, Mind e il linguaggio in Ludwig Wittgenstein e John Searle' 2nd ed (2019). Coloro che sono interessati a più dei miei scritti possono vedere 'TalkingMonkeys--Filosofia, Psicologia, Scienza, Religione e Politica su un Pianeta Condannato--Articoli e Recensioni 2006-2019 3rd ed (2019) e Suicidal Utopian Delusions nel 21st Century 4th ed (2019) . (shrink)

This paper proposes that existing computational modeling research programs may be combined into platforms for the information of public policy. The main idea is that computational models at select levels of organization may be integrated in natural terms describing biological cognition, thereby normalizing a platform for predictive simulations able to account for both human and environmental costs associated with different action plans and institutional arrangements over short and long time spans while minimizing computational requirements. Building from established research programs, the (...) proposal aims to take advantage of current momentum in the direction of the integration of the cognitive with social and natural sciences, reduce start-up costs and increase speed of development. These are all important upshots given rising unease over the potential for AI and related technologies to shape the world going forward. (shrink)

Le déficit d’explicabilité des techniques d’apprentissage machine (AM) pose des problèmes opérationnels, juridiques et éthiques. Un des principaux objectifs de notre projet est de fournir des explications éthiques des sorties générées par une application fondée sur de l’AM, considérée comme une boîte noire. La première étape de ce projet, présentée dans cet article, consiste à montrer que la validation de ces boîtes noires diffère épistémologiquement de celle mise en place dans le cadre d’une modélisation mathématique et causale d’un phénomène physique. (...) La différence majeure est qu’une méthode d’AM ne prétend pas représenter une causalité entre les paramètres d’entrées, qui peuvent être de plus de haute dimensionnalité, et ceux de sortie. Nous montrons dans cet article l’intérêt de mettre en œuvre les distinctions épistémologiques entre les différentes fonctions épistémiques d’un modèle, d’une part, et entre la fonction épistémique et l’usage d’un modèle, d’autre part. Enfin, la dernière partie de cet article présente nos travaux en cours sur l’évaluation d’une explication, qui peut être plus persuasive qu’informative, ce qui peut ainsi causer des problèmes d’ordre éthique. (shrink)

In 2017 AlphaZero, a neural network-based chess engine shook the chess world by convincingly beating Stockfish, the highest-rated chess engine. In this paper, I describe the technical differences between the two chess engines and based on that, I discuss the impact of the modeling choices on the respective epistemic opacities. I argue that the success of AlphaZero’s approach with neural networks and reinforcement learning is counterbalanced by an increase in the epistemic opacity of the resulting model.

According to the forecast that billions of devices will get connected to the Internet by 2020. All these devices will produce a huge amount of data that will have to be handled rapidly and in a feasible manner. It will become a challenge for real-time applications to handle this huge data while considering security issues as well as time constraints. The main highlights of cloud computing are on-demand service and scalability; therefore the data generated from IoT devices are generally handled (...) in cloud infrastructure. Though, dealing with IoT application requests on the cloud exclusively is not a proficient result for some IoT applications particularly time-sensitive ones. These issues can be settled by utilizing another idea called, Fog computing. Fog computing has become one of the major fields of research from both academia and industry perspectives. The ongoing research commitments on few issues in fog computing are figuring out in this paper. At long last, this paper also highlights some open issues in fog with IoT, which will determine the future research direction for implementing Fog computing paradigm. (shrink)

In the last few years, biologists and computer scientists have claimed that the introduction of data science techniques in molecular biology has changed the characteristics and the aims of typical outputs (i.e. models) of such a discipline. In this paper we will critically examine this claim. First, we identify the received view on models and their aims in molecular biology. Models in molecular biology are mechanistic and explanatory. Next, we identify the scope and aims of data science (machine learning in (...) particular). These lie mainly in the creation of predictive models which performances increase as data set increases. Next, we will identify a tradeoff between predictive and explanatory performances by comparing the features of mechanistic and predictive models. Finally, we show how this a priori analysis of machine learning and mechanistic research applies to actual biological practice. This will be done by analyzing the publications of a consortium—The Cancer Genome Atlas—which stands at the forefront in integrating data science and molecular biology. The result will be that biologists have to deal with the tradeoff between explaining and predicting that we have identified, and hence the explanatory force of the ‘new’ biology is substantially diminished if compared to the ‘old’ biology. However, this aspect also emphasizes the existence of other research goals which make predictive force independent from explanation. (shrink)

Many of the computing systems programmed using Machine Learning are opaque: it is difficult to know why they do what they do or how they work. Explainable Artificial Intelligence aims to develop analytic techniques that render opaque computing systems transparent, but lacks a normative framework with which to evaluate these techniques’ explanatory successes. The aim of the present discussion is to develop such a framework, paying particular attention to different stakeholders’ distinct explanatory requirements. Building on an analysis of “opacity” from (...) philosophy of science, this framework is modeled after accounts of explanation in cognitive science. The framework distinguishes between the explanation-seeking questions that are likely to be asked by different stakeholders, and specifies the general ways in which these questions should be answered so as to allow these stakeholders to perform their roles in the Machine Learning ecosystem. By applying the normative framework to recently developed techniques such as input heatmapping, feature-detector visualization, and diagnostic classification, it is possible to determine whether and to what extent techniques from Explainable Artificial Intelligence can be used to render opaque computing systems transparent and, thus, whether they can be used to solve the Black Box Problem. (shrink)

We present nine facets for the analysis of the past and future evolution of AI. Each facet has also a set of edges that can summarise different trends and contours in AI. With them, we first conduct a quantitative analysis using the information from two decades of AAAI/IJCAI conferences and around 50 years of documents from AI topics, an official database from the AAAI, illustrated by several plots. We then perform a qualitative analysis using the facets and edges, locating AI (...) systems in the intelligence landscape and the discipline as a whole. This analytical framework provides a more structured and systematic way of looking at the shape and boundaries of AI. (shrink)

Die Entwicklungen in der Künstlichen Intelligenz (KI) sind spannend. Aber wohin geht die Reise? Ich stelle eine Analyse vor, der zufolge exponentielles Wachstum von Rechengeschwindigkeit und Daten die entscheidenden Faktoren im bisherigen Fortschritt waren. Im Folgenden erläutere ich, unter welchen Annahmen dieses Wachstum auch weiterhin Fortschritt ermöglichen wird: 1) Intelligenz ist eindimensional und messbar, 2) Kognitionswissenschaft wird für KI nicht benötigt, 3) Berechnung (computation) ist hinreichend für Kognition, 4) Gegenwärtige Techniken und Architektur sind ausreichend skalierbar, 5) Technological Readiness Levels (TRL) (...) erweisen sich als machbar. Diese Annahmen werden sich als dubios erweisen. (shrink)

Cloud Computing paradigm has reached a high degree of popularity among all kinds of computer users, but it may not be suitable for mobile devices as they need computing power to be as close as possible to data sources in order to reduce delays. This paper focuses on achieving mathematical models for users moving around and proposes an overlay mobility model for Fog Data Centres based on traditional wireless mobility models aimed at better allocating edge computing resources to client demands. (...) Additionally, the focus is also on providing a model as to how the computing assets related to each user follow it through the fog infrastructure so as to minimise the distance to the user according to a Fat-Tree based Data Centre topology. (shrink)

Engineers fine-tune the design of robot bodies for control purposes, however, a methodology or set of tools is largely absent, and optimization of morphology (shape, material properties of robot bodies, etc.) is lagging behind the development of controllers. This has become even more prominent with the advent of compliant, deformable or ”soft” bodies. These carry substantial potential regarding their exploitation for control—sometimes referred to as ”morphological computation”. In this article, we briefly review different notions of computation by physical systems and (...) propose the dynamical systems framework as the most useful in the context of describing and eventually designing the interactions of controllers and bodies. Then, we look at the pros and cons of simple vs. complex bodies, critically reviewing the attractive notion of ”soft” bodies automatically taking over control tasks. We address another key dimension of the design space—whether model-based control should be used and to what extent it is feasible to develop faithful models for different morphologies. (shrink)

Despite the impressive amount of financial resources recently invested in carrying out large-scale brain simulations, it is controversial what the pay-offs are of pursuing this project. One idea is that from designing, building, and running a large-scale neural simulation, scientists acquire knowledge about the computational performance of the simulating system, rather than about the neurobiological system represented in the simulation. It has been claimed that this knowledge may usher in a new era of neuromorphic, cognitive computing systems. This study elucidates (...) this claim and argues that the main challenge this era is facing is not the lack of biological realism. The challenge lies in identifying general neurocomputational principles for the design of artificial systems, which could display the robust flexibility characteristic of biological intelligence. (shrink)

We overview the main historical and technological elements characterising the rise, the fall and the recent renaissance of the cognitive approaches to Artificial Intelligence and provide some insights and suggestions about the future directions and challenges that, in our opinion, this discipline needs to face in the next years.

A lot of infants have significant food-related problems, as well as spitting up, rejecting new foods, or not accepting to eat at specific times. These issues are frequently ordinary and are not a sign that the baby is unwell. According to the National Institutes of Health, 25% of generally developing infants and 35% of babies with neurodevelopmental disabilities are tormented by some sort of feeding problem. Some, for example rejecting to eat specific foods or being overly finicky, are momentary and (...) don’t cause any health dangers. This paper proposes an expert system that can be used to successfully diagnose Feeding problems in infants and children. The suggested systems were found to be beneficial approach in addition to existing impartial ones. So far as the authors are aware, this is the initial effort of using an expert system in attaining good performance in a real world application. This expert system was designed and implemented to help parents diagnose these problems and get a recommendation of how to deal with infants and children. (shrink)

Despite the impressive amount of financial resources invested in carrying out large-scale brain simulations, it is controversial what the payoffs are of pursuing this project. The present paper argues that in some cases, from designing, building, and running a large-scale neural simulation, scientists acquire useful knowledge about the computational performance of the simulating system, rather than about the neurobiological system represented in the simulation. What this means, why it is not a trivial lesson, and how it advances the literature on (...) the epistemology of computer simulation are the three preoccupations addressed by the paper. (shrink)

What is the class of possible semiotic systems? What kinds of systems could count as such systems? The human mind is naturally considered the prototypical semiotic system. During years of research in semiotics the class has been broadened to include i.e. living systems like animals, or even plants. It is suggested in the literature on artificial intelligence that artificial agents are typical examples of symbol-processing entities. It also seems that semiotic processes are in fact cognitive processes. In consequence, it is (...) natural to ask the question about the relation between semiotic studies and research on artificial cognitive systems within cognitive science. Consequently, my main question concerns the problem of inclusion or exclusion from the semiotic spectrum at least some artificial systems. I would like to consider some arguments against the possibility of artificial semiotic systems and I will try to repeal them. Then I will present an existing natural-language using agent of the SNePS system and interpret it in terms of Peircean theory of signs. I would like also to show that some properties of semiotic systems in Peircean sense could be also found in a discussed artificial system. Finally, I will have some remarks on the status of semiotics in general. (shrink)

Artificial models of cognition serve different purposes, and their use determines the way they should be evaluated. There are also models that do not represent any particular biological agents, and there is controversy as to how they should be assessed. At the same time, modelers do evaluate such models as better or worse. There is also a widespread tendency to call for publicly available standards of replicability and benchmarking for such models. In this paper, I argue that proper evaluation ofmodels (...) does not depend on whether they target real biological agents or not; instead, the standards of evaluation depend on the use of models rather than on the reality of their targets. I discuss how models are validated depending on their use and argue that all-encompassing benchmarks for models may be well beyond reach. (shrink)

We present an approach to subjective computing for the design of future robots that exhibit more adaptive and flexible behavior in terms of subjective intelligence. Instead of encapsulating subjectivity into higher order states, we show by means of a relational approach how subjective intelligence can be implemented in terms of the reciprocity of autonomous self-referentiality and direct world-coupling. Subjectivity concerns the relational arrangement of an agent’s cognitive space. This theoretical concept is narrowed down to the problem of coaching a reinforcement (...) learning agent by means of binary feedback. Algorithms are presented that implement subjective computing. The relational characteristic of subjectivity is further confirmed by a questionnaire on human perception of robot’s behavior. The results imply that subjective intelligence cannot be externally observed. In sum, we conclude that subjective intelligence in relational terms is fully tractable and therefore implementable in artificial agents. (shrink)

A Bayesian network (BN) is a graphical model of uncertainty that is especially well suited to legal arguments. It enables us to visualize and model dependencies between different hypotheses and pieces of evidence and to calculate the revised probability beliefs about all uncertain factors when any piece of new evidence is presented. Although BNs have been widely discussed and recently used in the context of legal arguments, there is no systematic, repeatable method for modeling legal arguments as BNs. Hence, where (...) BNs have been used in the legal context, they are presented as completed pieces of work, with no insights into the reasoning and working that must have gone into their construction. This means the process of building BNs for legal arguments is ad hoc, with little possibility for learning and process improvement. This article directly addresses this problem by describing a method for building useful legal arguments in a consistent and repeatable way. The method complements and extends recent work by Hepler, Dawid, and Leucari (2007) on object-oriented BNs for complex legal arguments and is based on the recognition that such arguments can be built up from a small number of basic causal structures (referred to as idioms). We present a number of examples that demonstrate the practicality and usefulness of the method. (shrink)

Concept representation is still an open problem in the field of ontology engineering and, more generally, of knowledge representation. In particular, the issue of representing “non classical” concepts, i.e. concepts that cannot be defined in terms of necessary and sufficient conditions, remains unresolved. In this paper we review empirical evidence from cognitive psychology, according to which concept representation is not a unitary phenomenon. On this basis, we sketch some proposals for concept representation, taking into account suggestions from psychological research. In (...) particular, it seems that human beings employ both prototype-based and exemplar-based representations in order to represent non classical concepts. We suggest that a similar, hybrid prototype-exemplar based approach could also prove useful in the field of knowledge representation technology. Finally, we propose conceptual spaces as a suitable framework for developing some aspects of this proposal. (shrink)

In this paper, we model a relational notion of subjectivity by means of two experiments in subjective computing. The goal is to determine to what extent a cognitive and social robot can be regarded to act subjectively. The system was implemented as a reinforcement learning agent with a coaching function. To analyze the robotic agent we used the method of levels of abstraction in order to analyze the agent at four levels of abstraction. At one level the agent is described (...) in mentalistic or subjective language respectively. By mapping this mentalistic to an algorithmic, functional, and relational level, we can show to what extent the agent behaves subjectively as we make use of a relational concept of subjectivity that draws upon the relations that hold between the agent and its environment. According to a relational notion of subjectivity, an agent is supposed to be subjective if it exhibits autonomous relations to itself and others, i.e. the agent is not fully determined by a given input but is able to operate on its input and decide what to do with it. This theoretical notion is confirmed by the technical implementation of self-referentiality and social interaction in that the agent shows improved behavior compared to agents without the ability of subjective computing. On the one hand, a relational concept of subjectivity is confirmed, whereas on the other hand, the technical framework of subjective computing is being theoretically founded. (shrink)