ChapterPDF Available

Getting Clarity by Defining Artificial Intelligence—A Survey

Authors:

Abstract and Figures

Intelligence remains ill-defined. Theories of intelligence and the goal of Artificial Intelligence (A.I.) have been the source of much confusion both within the field and among the general public. Studies that contribute to a well-defined goal of the discipline and spread a stronger, more coherent message, to the mainstream media, policy-makers, investors, and the general public to help dispel myths about A.I. are needed. We present the preliminary results of our research survey “Defining (machine) Intelligence.” Opinions, from a cross sector of professionals, to help create a unified message on the goal and definition of A.I.
Content may be subject to copyright.
Getting Clarity by Defining Artificial
Intelligence—A Survey
Dagmar Monett1,2(B
)and Colin W. P. Lewis1(B
)
1AGISI.org, Berlin, Germany
{dagmar.monett,colin.lewis}@agisi.org
2HWR Berlin, Berlin, Germany
Abstract. Intelligence remains ill-defined. Theories of intelligence and
the goal of Artificial Intelligence (A.I.) have been the source of much
confusion both within the field and among the general public. Stud-
ies that contribute to a well-defined goal of the discipline and spread
a stronger, more coherent message, to the mainstream media, policy-
makers, investors, and the general public to help dispel myths about
A.I. are needed. We present the preliminary results of our research sur-
vey “Defining (machine) Intelligence.” Opinions, from a cross sector of
professionals, to help create a unified message on the goal and definition
of A.I.
1 Introduction
Intelligence permeates almost everything we do. Formally providing a robust and
scientific definition of intelligence has been a goal of scientists and researchers for
several centuries. During the last sixty years, the formal definition of intelligence
has taken on extra impetus as machine intelligence, or A.I., developers pursue
their vision of creating intelligent machines that “replicate” human intelligence
(Brooks 1991). For others, the goal is creating Artificial General Intelligent sys-
tems which exceed human intelligence. However, it is still very hard to define
what intelligence is (Kambhampati 2017). Furthermore, creating an agreed upon
message on the goal and definition of A.I. is far from obvious or straightforward
(Nilsson 2010).
In order to clarify the goal and definition of A.I. the research survey “Defining
(machine) Intelligence” solicits opinions from a cross sector of professionals. The
ongoing survey1has attracted a significant volume of responses and high level
comments and recommendations concerning the definitions of A.I. and human
intelligence from experts around the world. We believe that collecting experts’
opinions can contribute to both a deeper understanding and a better definition
of what intelligence is. In this short paper, a partial analysis of the first 400
responses is presented.
1See http://agisi.org/Survey intelligence.html for more about the survey.
c
Springer Nature Switzerland AG 2018
V. C. M¨uller (Ed.): PT-AI 2017, SAPERE 44, pp. 212–214, 2018.
https://doi.org/10.1007/978-3-319-96448-5_21
Getting Clarity by Defining AI 213
2 Research Survey: Preliminary Results and Discussion
The research survey focuses on specific definitions of human and machine
intelligence,2and on the level of agreement of respondents with those defini-
tions. The survey further asks respondents to provide their level of agreement
with statements based on DeBoeck’s (2013) questions concerning the defini-
tion of intelligence. Potential respondents were collected from different sources
with research topics relevant to the survey. News lists informing computer sci-
entists, neuroscientists, cognitive sociologists, and roboticists, to name a few,
were also considered. Survey respondents originate from 48 countries and 131
different institutions (academia 77%, industry 21.3%). Respondents are mainly
researchers (75.3%), educators (36%), and developers or engineers (16.8%), and
come from Computer Science (58%), Psychology or Cognitive Science (9.3%),
and Engineering (8.5%).
Partial results show (see Table 1) that most respondents disagree or strongly
disagree there is a difficulty in defining the goal of A.I. (1.b) Only a small minor-
ity seem to believe that a definition of intelligence is self-evident (1.a) and over
the half of respondents disagree that it will never be possible to reach an agree-
ment upon a definition of A.I. (1.f) Most respondents indicated disagreement
or strong disagreement with the statement that a unified definition of artificial
intelligence does not pay off (1.d). However, there are also strong opinions for the
contrary and an almost equal amount of opinions are neutral. Other statements
to agree upon considered differences in opinion when defining A.I. being too
large to bridge (1.c), a definition of A.I. experienced as a restriction (1.e), and
scientific advances in A.I. being “a huge step forward and possibly a promising
paradigm shift towards creating machines that can be measured to match or
exceed human level intelligence” (1.g).
Table 1. Level of agreement with some statements to agree upon in the survey
(N= 400).
Id Strongly disagree Disagree Neutral Agree Strongly agree
1.a 160 (40.0%) 169 (42.3%) 34 (8.5%) 24 (6.0%) 13 (3.3%)
1.b 61 (15.3%) 173 (43.3%) 83 (20.8%) 71 (17.8%) 12 (3.0%)
1.c 19 (4.8%) 152 (38.0%) 116 (29.0%) 93 (23.3%) 20 (5.0%)
1.d 26 (6.5%) 125 (31.3%) 119 (29.8%) 115 (28.8%) 15 (3.8%)
1.e 51 (12.8%) 160 (40.0%) 81 (20.3%) 93 (23.3%) 15 (3.8%)
1.f 55 (13.8%) 171 (42.8%) 94 (23.5%) 62 (15.5%) 18 (4.5%)
1.g 40 (10.0%) 66 (16.5%) 74 (18.5%) 147 (36.8%) 73 (18.3%)
Many respondents (N= 187, 46.8%) express agreement or strong agree-
ment concerning the need for having separate definitions of human and machine
2See http://agisi.org/Defs intelligence.html for a complete list of definitions.
214 D. Monett and C. W. P. Lewis
intelligence, but a slightly equal number (N= 172, 43%) indicate that only
one definition is adequate. The definition of machine intelligence that received
the most comments was Russell and Norvig’s (2010) definition with a total of
224 (56%) opinions. However, the most accepted definition of machine intelli-
gence was Wang’s (2008): 224 (56%) respondents agree or strongly agree with it.
Similarly, the definition of human intelligence that received the most comments
was Humphreys’ (1984) definition with a total of 148 (37%) opinions. However,
the most accepted definition was Gottfredson’s (1997) definition: 246 (61.5%)
respondents agree or strongly agree with it. Survey participants provided a total
of 3453 reasons for supporting their selections of the definitions. Furthermore, a
total of 213 (53.3%) survey participants provided their suggested definitions of
human and/or machine intelligence.3
3 Conclusions
Getting clarity around defining A.I. must include experts’ opinions. The first 400
responses to our survey comprise thousands of those opinions, not to mention
other hundreds of suggested definitions of intelligence and overall feedback that
were received. A significant variety of judgements and viewpoints that allow for
first understanding and then creating an agreed upon message on the goal and
definition of A.I. The question of combined versus separate definitions of human
and machine intelligence remains highly polarized. Our work in progress includes
building a catalogue of factors contributing to intelligence, and a methodology
for and best practices to be applied when defining (machine) intelligence.
References
Brooks, R.A.: Intelligence without representation. Artif. Intell. 47, 139–159 (1991)
De Boeck, P.: Intelligence, where to look, where to go? J. Intell. 1, 5–24 (2013)
Gottfredson, L.S.: Mainstream science on intelligence: an editorial with 52 signatories,
history, and bibliography. Intelligence 24, 13–23 (1997)
Humphreys, L.G.: General intelligence. In: Reynolds, C.R., Brown, R.T. (eds.)
Perspectives on Bias in Mental Testing, p. 243. Springer, Boston (1984)
Kambhampati, S.: On the Past and Future of AI. Interviews with Experts in Artificial
Intelligence, Iridescent (2017). https://goo.gl/nspv6y
Nilsson, N.J.: The Quest for Artificial Intelligence. A History of Ideas and Achieve-
ments. Cambridge University Press, Cambridge (2010)
Russell, S.J., Norvig, P.: Artificial Intelligence: A Modern Approach, 3rd edn. Prentice
Hall, Upper Saddle River (2010)
Wang, P.: What do you mean by “AI”? In: Wang, P., Goertzel, B., Franklin, S. (eds.)
Proceedings of the First AGI Conference on Artificial General Intelligence 2008.
Frontiers in Artificial Intelligence and Applications, vol. 171, pp. 362-373. IOS Press,
Amsterdam (2008)
3More results can be found in our poster to the PT-AI 2017 conference (Leeds, UK)
that is available at http://agisi.org/doc/Monett-Lewis PT- AI2017 poster.pdf.
... Because the concept of AI from a cross-section of experts was still ill-defined (Monett & Lewis, 2018), literature on AI philosophy in the areas of philosophy, particularly philosophy of sciences like biology and physics, remained fragmented. Others explored ethical difficulties in the practice of AI researchers and developers, while others designated a new idea of intelligence, while others commented on designs and the relationships between AI and humans. ...
... The search for a common definition of AI is still ongoing (Monett & Lewis, 2018). While AI engineers work towards their goal of building intelligence robots with human-level intellect (Brooks, 1991;McCarthy, 2015), defining what types of computing methods are called intelligence remains difficult (Kambhampati, 2017). ...
... In a review of current and teaching of AI, Wollowski et al. (2016) discovered that a definition of AI for educators and practitioners was still far from being established due to the educators' varied opinions on historical, ethical, and philosophical issues. The necessity for defining established definitions of intelligence and AI research goals was identified as an effort by AGISI.org, which began a project in 2018 to find a clear definition of human and machine intelligence based on expert viewpoints (Monett & Lewis, 2018). ...
Article
Full-text available
Background In fact, most schools around the world are not well equipped to have discussions and keep current on the expansion of artificial intelligence (AI) in many aspects of society and economy. They either ignore this conversation, or simply criticize technology, but these resistances are not stopping wide spread of various types of AI projects in schools, mainly driven by corporations, and fueled by incentives that might not match well with long term educational objectives of student success, diversity, equity, and inclusivity. Objectives The overall purpose is to address the rising gap between the ultrafast development of AI and the meticulous technological application of education, and to suggest the important bridge of building technological leadership in teacher preparation to get ready for the grow of AI in education. Methods The contextual review examines the field of AI development in education through three lenses: conceptual context, practice context, and research context. Results and Conclusions The paper provides educators and policy makers an overall background of the phenomenon of AI in education. The study has revealed there is an urgent need for research and development in teacher preparation as well as in the philosophy of technology in education to bridge the gap between AI and education.
... The annotation corpus consists of four collections of definitions of intelligence. Participants to a survey on definitions of intelligence [17] were asked to provide their level of agreement with definitions of both human and machine intelligence from the literature (for more on the survey, please consult the provided reference). They were also asked to justify their selection, as well as to provide new definitions of intelligence, if desired. ...
... A New, suggested definitions of machine or artificial intelligence by participants to the survey on defining intelligence [17]. ...
... New, suggested definitions of human intelligence by participants to the survey on defining intelligence [17]. ...
Conference Paper
Full-text available
Delineating the boundaries of the discourse on machine or artificial intelligence (AI) may help in defining and understanding its most discussed concept, the concept of intelligence. Furthermore, better insights into both definitions and how to define them well has proven to be essential for a better understanding of concepts, intelligence included. These and related cognitive abilities (e.g. defining, analyzing, understanding, discussing, and comparing definitions of intelligence, among others) are expected for AI researchers and practitioners in the first place. Yet, they are also central to extending or at least providing the basics of AI literacy to other stakeholders of our society. Intelligent systems are transforming the way we interact with technology, with each other, and with ourselves, and knowing at least what AI or intelligence mean is becoming essential for designing, developing, deploying, using, and even regulating intelligent artefacts. However, defining intelligence has been one of the most controversial and studied challenges of both ancient and modern human thinking. A lack of consensus on what intelligence is has remained almost constant over the centuries. Interested scholars have not come up with a consensus or cross-domain accepted definition of intelligence. Neither in the ancient Eastern nor in the ancient and contemporary Western conceptions of intelligence. Nor in the more recent perspectives from the last 70 years within the field of AI. We argue that a better understanding of contemporary technologies, AI-based but not only, starts with a grounded exposure to their conceptual pillars. These include fundamental concepts like the concept of intelligence, in general, and of AI, in particular. Learners and decision makers at all levels should face them, as well as be able to discuss their importance and limitations critically and in an informed way. For doing that, they must be confronted with definitions of intelligence and understand their meaning well, for instance. If these contents are already part of study programs, the better. It is the main goal of this paper to present how a few hundreds of definitions of intelligence were annotated, i.e. their properties and characteristics systematically analyzed and commented, in order to construct a corpus (i.e. a collection) of definitions of intelligence for further uses in AI and other fields. The work and particular application domain presented here has not yet been considered in the extended work on linguistic annotation. Even though, both the annotation and the data merit special attention, for they deal with the elusive, important concept of intelligence, i.e. with definitions of both human and machine (or artificial) intelligence. Undergraduate Computer Science students carried out the annotation process and several research activities. They were involved in an AI research project led by faculty and included their findings and work as part of their undergraduate student research projects in their last study year. We provide details about how the student research projects were conceived, conducted, and mentored. We also describe the properties or quality criteria that were considered for annotating the definitions from the intelligence corpus.
... With the term being coined by John McCarthy during a summer conference at Dartmouth College in 1956 [23], roots of AI started. However, at present there is no universally accepted scholarly definition of AI [24], with its status quo being described as the parable of "the blind men and an elephant" [26]. ...
... Perhaps the most significant challenge to establish a well-accepted definition is that AI relates to different branches of the knowledge domain due to its historical development [39]. Because of the varied perspectives of each AI research field, answers to the questions of what constitutes AI and what is appropriate terminology have been inconsistent [24]. The relentless pace of technological advancement further complicates the definition of AI [16]. ...
Chapter
This paper reports on a scoping review of research studies on artificial intelligence in education (AIED) published over the last two decades (2001–2021). A wide range of manuscripts were yielded from the education and educational research category of the Social Sciences Citation Index (SSCI) database, and papers from an AIED-specialised journal were also included. 135 of those meeting the selection criteria were analysed with content analysis and categorical meta-trends analysis. Three distinctive and superordinate AIED research agenda were identified: Learning from AI, Learning about AI, and Learning with AI. By portraying the current status of AIED research and depicting its tripartite research agenda, gaps and possible directions were discussed. This paper serves as a blueprint for AIED researchers to position their up-and-coming AIED studies for the next decade.
... How does one define AI? According to a survey conducted by Artificial General Intelligence Sentinel Initiative in 2018, the most agreeable definition for AI voted by experts is that stated in [1]. There, AI is described as having the faculty of adaptation and improvisation, (i) Failures of robustness: The system is subjected to unusual or unforeseen inputs, causing failures. ...
... consists of a non-parametric baseline, a linear component of the current condition, and a recurrent component as proposed in [46] for turbofan prognostic application, with the model being represented in (1). Here, the first component, λ(t), is the non-parametric part consisting of a lifetime probabilistic model. ...
Article
Full-text available
Mistrust, amplified by numerous artificial intelligence (AI) related incidents, is an issue that has caused the energy and industrial sectors to be amongst the slowest adopter of AI methods. Central to this issue is the black-box problem of AI, which impedes investments and is fast becoming a legal hazard for users. Explainable AI (XAI) is a recent paradigm to tackle such an issue. Being the backbone of the industry, the prognostic and health management (PHM) domain has recently been introduced into XAI. However, many deficiencies, particularly the lack of explanation assessment methods and uncertainty quantification, plague this young domain. In the present paper, we elaborate a framework on explainable anomaly detection and failure prognostic employing a Bayesian deep learning model and Shapley additive explanations (SHAP) to generate local and global explanations from the PHM tasks. An uncertainty measure of the Bayesian model is utilized as a marker for anomalies and expands the prognostic explanation scope to include the model’s confidence. In addition, the global explanation is used to improve prognostic performance, an aspect neglected from the handful of studies on PHM-XAI. The quality of the explanation is examined employing local accuracy and consistency properties. The elaborated framework is tested on real-world gas turbine anomalies and synthetic turbofan failure prediction data. Seven out of eight of the tested anomalies were successfully identified. Additionally, the prognostic outcome showed a 19% improvement in statistical terms and achieved the highest prognostic score amongst best published results on the topic.
... Consequently, many definitions of machine intelligence exist (Monett and Lewis 2017), which tend to be influenced by different theoretical backgrounds and practical considerations. Most definitions of machine intelligence consider it a demonstration, or an effect, of specific capabilities of an agent, which ''encompass at least the essence of human intelligence along with the prospect of other capabilities specific to machines'' (Legg and Hutter 2008). ...
Article
Full-text available
Intelligent assistants are an increasingly commonplace class of information systems spanning a broad range of form and complexity. But what characterizes an intelligent assistant, and how do we design better assistants? In the paper, the authors contribute to scientific research in the domain of intelligent assistants in three steps, each building on the previous. First, they investigate the historical context of assistance as human work. By examining qualitative studies regarding the work of human assistants, the authors inductively derive concepts crucial to modeling the context of assistance. This analysis informs the second step, in which they develop a conceptual typology of intelligent assistants using 111 published articles. This typology explicates the characteristics (what or how) of intelligent assistants and their use context (who or which). In the third and final step, the authors utilize this typology to shed light on historical trends and patterns in design and evaluation of intelligent assistants, reflect on missed opportunities, and discuss avenues for further exploration.
... Together with statistics, mathematics, data mining, and computer science they form part of a wider domain called data science [20] AI is at the core of the transformational impacts that digital technologies are having in human lives and infrastructure. However, there is no common standard definition for AI [21][22][23][24]. The Oxford Dictionary [25] defines it as "The theory and development of computer systems able to perform tasks normally requiring human intelligence, such as visual perception, speech recognition, decision-making, and translation between languages". ...
Article
As Building Information Modeling (BIM) workflows are becoming very relevant for the different stages of the project’s lifecycle, more data is produced and managed across it. The information and data accumulated in BIM-based projects present an opportunity for analysis and extraction of project knowledge from the inception to the operation phase. In other industries, Machine Learning (ML) has been demonstrated to be an effective approach to automate processes and extract useful insights from different types and sources of data. The rapid development of ML applications, the growing generation of BIM-related data in projects, and the different needs for use of this data present serious challenges to adopt and effectively apply ML techniques to BIM-based projects in the Architecture, Engineering, Construction and Operations (AECO) industry. While research on the use of BIM data through ML has increased in the past decade, it is still in a nascent stage. In order to asses where the industry stands today, this paper carries out a systematic literature review (SLR) identifying and summarizing common emerging areas of application and utilization of ML within the context of BIM-generated data. Moreover, the paper identifies research gaps and trends. Based on the observed limitations, prominent future research directions are suggested, focusing on information architecture and data, applications scalability, and human information interactions.
Technical Report
Full-text available
This research report has been written as part of the Kajaani University of Applied Sciences (KAMK) AI Boost project (ELY funding for the development of companies’ operative environment in the European Regional Development Fund). AI Boost is included as one of the CEMIS (Centre of Measurement and Information Systems, www.cemis.fi) projects between 2020-2021. The AI Boost project aims to create new knowhow and understanding on how companies in Finland are utilizing artificial intelligence (AI) in their activities. The project aims to clarify the drivers and challenges related to deploying AI for small and medium-sized companies. In addition, the pur�pose is to identify how AI utilization can impact company strategy, turnover, and results. Addi�tionally, there is the target of understanding how companies can create new business, increase current business, and make operations more efficient, through AI.
Article
Full-text available
Diante da importância atribuída à inteligência artificial (IA) e devido ao seu caráter transversal e pervasivo, muitos governos têm elaborado estratégias nacionais de IA, em um movimento de competição tecnológica. Até o momento, 56 países, além da União Europeia (UE), possuem documentos mais ou menos sistematizados, os quais foram mapeados pelo Observatório de Políticas de IA da Organização para a Cooperação e o Desenvolvimento Econômico (OCDE). Embora existam estudos que proponham análises destas estratégias sob diferentes perspectivas e metodologias, há ainda espaço para debatê-las a partir de outros pontos de vista, sobretudo com base na consistência do diagnóstico dos problemas-chave identificados em cada um dos documentos. Os países examinados neste Texto para Discussão são Argentina, Brasil, Colômbia, Chile e Coreia do Sul. As estratégias do grupo sul-americano e da Coreia do Sul não estão isentas de lacunas em relação ao diagnóstico dos problemas-chave: na maioria das vezes, as estratégias apresentaram diagnósticos baseados em evidências insuficientes e pouco consistentes, usando com frequência indicadores inadequados e até mesmo insatisfatórios para subsidiá-los. As análises qualitativas apresentadas trazem uma contribuição dupla para os debates sobre estratégias nacionais de IA. Em primeiro lugar, elas dão luz ao Estado como ator central em um campo dominado por arranjos de governança privada, revelando a possibilidade das interações híbridas em formação. Em segundo lugar, elas capturam a variedade de abordagens adotadas pelos governos para responder aos desafios da IA, embora as estratégias de Argentina, Brasil, Chile e Colômbia estejam atreladas ao movimento global puxado por outros países.
Chapter
Full-text available
"Industrial revolution 4.0" is a term that is becoming increasingly popular among academics. A number of articles have been carved to emphasize the beneficial aspects of the stated issue under many titles such as cyber physical systems, internet of things, artificial intelligence, smart manufacturing, and digitalization of industrial production, and so on. However, few academics have delved into the negative or dark side of such a profound technological paradigm change, especially the artificially intelligent robotics, creating a large knowledge vacuum. Because of this, little is known about the negative repercussions of artificial intelligence (AI), a key component of the Fourth Industrial Revolution (or IR 4.0). It is an open secret now that AI machines may have serious impacts on human autonomy, fairness, justice and agency. These unanticipated consequences have resulted in the development of an emerging concept, i.e., responsible innovation. The responsible innovation framework binds the firm ethically, morally, and socially to be responsible, environmentally friendly, humanitarian, and business-oriented while developing innovative products. The current study proposes an integrated responsible innovation framework that acts as a science governance mechanism and considers organizations and stakeholders collectively responsible for upcoming technological innovations. This study has suggested several implications for policymakers.
Article
Full-text available
The first issue of the Journal of Intelligence is devoted to a discussion based on the following two questions: 1. What are the most important scientific issues in the domain of human intelligence? 2. What are the most promising new ideas and approaches in the study of human intelligence? [...]
Conference Paper
Full-text available
Many problems in AI study can be traced back to the confusion of different research goals. In this paper, five typical ways to define AI are clarified, analyzed, and compared. It is argued that though they are all legitimate research goals, they lead the research to very different directions, and most of them have trouble to give AI a proper identity. Finally, a working definition of AI is proposed, which has important advantages over the alternatives.
Article
JIT is not a single variable problem. It is not just inventory reduction. A more useful definition begins with cost curves. The carrying cost of the materials in a manufacturing operation is shown to increase with the amount of inventory on hand. But the cost of production disruption decreases with inventory - the more material you have, the less effect a machine breakdown will have. The sum of these two cost curves results in a total cost curve. The minimum point on the total cost curve indicates the optimal trade-off between inventory carrying costs and production disruption costs. Merely reducing inventory without taking other measures would actually increase the total cost. To move the minimum point down, the carrying cost curve or the disruption cost curve (or both) have be shifted downward. An accompanying effect is then a reduction in inventory levels.
Article
This statement outlines conclusions re-garded as mainstream among researchers on intelligence, in particular, on the nature, ori-gins, and practical consequences of individu-al and group differences in intelligence. Its aim is to promote more reasoned discussion of the vexing phenomenon that the research has revealed in recent decades. The follow-ing conclusions are fully described in the major textbooks, professional journals and encyclopedias in intelligence.
Chapter
Both the theory of intelligence and its measurement have been confused by failure to follow a scientifically acceptable definition of intelligence. In good part, the problem arises from the folk definition of intelligence that involves the innate fixed capacity of a person to solve problems. This folk definition differs in only one important respect in its usage by environmentalists as opposed to hereditarians. The former tend to minimize the importance of individual differences in this fixed capacity, holding that any normal person has the fundamental capacity to achieve at a high level. Environmentalists are also strongly convinced that intelligence tests do not measure this capacity.
Book
Artificial intelligence (AI) is a field within computer science that is attempting to build enhanced intelligence into computer systems. This book traces the history of the subject, from the early dreams of eighteenth-century (and earlier) pioneers to the more successful work of today’s AI engineers. AI is becoming more and more a part of everyone’s life. The technology is already embedded in face-recognizing cameras, speech-recognition software, Internet search engines, and health-care robots, among other applications. The book’s many diagrams and easy-to-understand descriptions of AI programs will help the casual reader gain an understanding of how these and other AI systems actually work. Its thorough (but unobtrusive) end-of-chapter notes containing citations to important source materials will be of great use to AI scholars and researchers. This book promises to be the definitive history of a field that has captivated the imaginations of scientists, philosophers, and writers for centuries.
Article
Artificial intelligence research has foundered on the issue of representation. When intelligence is approached in an incremental manner, with strict reliance on interfacing to the real world through perception and action, reliance on representation disappears. In this paper we outline our approach to incrementally building complete intelligent Creatures. The fundamental decomposition of the intelligent system is not into independent information processing units which must interface with each other via representations. Instead, the intelligent system is decomposed into independent and parallel activity producers which all interface directly to the world through perception and action, rather than interface to each other particularly much. The notions of central and peripheral systems evaporate—everything is both central and peripheral. Based on these principles we have built a very successful series of mobile robots which operate without supervision as Creatures in standard office environments.
On the Past and Future of AI
  • S Kambhampati
Kambhampati, S.: On the Past and Future of AI. Interviews with Experts in Artificial Intelligence, Iridescent (2017). https://goo.gl/nspv6y