Contexte gauche	Mot	Contexte droit
	ai
	ai
	ai
The Impact of Multiple Narrow	AI	Technology
Sec.	AI	for Human Learning and Behavior
	ai	.2021.622364
Human- versus	Artificial Intelligence
	AI	is one of the most
human intelligence and	artificial intelligence	. Discussions on many
	ai
as the golden standard for	Artificial Intelligence	. In order to provide
between human- and	artificial intelligence	: 1) the fundamental
	ai
narrow-hybrid	AI	applications. For the time being
	AI	systems will have
to leave to	AI	and when is human judgment
intelligence? How to deploy	AI	systems effectively to complement and
	ai
we pursue the development of	AI	“partners” with human
these questions, humans working with	AI
AI. So, in order to obt	ai	n
well-functioning human-	AI	systems
in information technology and in	AI	may allow for more
Human-Aware AI, which	ai
ms at	AI	that adapts as a “team
When human-aware	AI	partners operate like “human collaborators
these “	AI	partners,” or “team mates” have
matter how intelligent and autonomous	AI	agents become in
	ai
	ai
working with advanced	AI	systems, (e.g. in military
capacities of	AI	systems in relation to human
will become increasingly relevant when	AI	systems become more advanced
	ai
use or “collaborate with” advanced	AI	systems in the near and
With the application of	AI	systems with increasing autonomy more
	ai
	ai
and architecture of biological and	artificial intelligence
	ai
to narrow, restricted tasks (narrow	AI
	ai
research differs from the ordinary	AI	research by
	ai
intelligence of	AI	will contribute to more adequate
	ai
	ai
a bit misleading) to vigorously	ai
upcoming issues) of	AI	in the short- and mid
a field as dynamic as	AI
articulated in the term “	Artificial Intelligence	”, as if it were not
progress in the field of	artificial intelligence	is accompanied
do.” In line with this,	AI	is then defined as “the
director of	AI	and a spokesman in the
nature of both human and	artificial intelligence	. This is
effectively using multiple narrow	AI	’s.^1
	ai
	ai
notion in this respect among	AI	scientists is that
	ai
	ai
	ai
	ai
	ai
	ai
	ai
	ai
	ai
	ai
	ai
	ai
	ai
	ai
	ai
	ai
	ai
	ai
	ai
	ai
	ai
Fundamental Differences Between Biological and	Artificial Intelligence
general	AI	means that a machine will
when we will reach general	AI	, (e.g., Goertzel
that the development of AI is determined by the constr	ai	nt
between human and	artificial intelligence	(Bostrom, 2014
contrast, if an AI system has learned a cert	ai	n
Speed: Signals from	AI	systems propagate with almost the
communication of	AI	systems that can be connected
Updatability and scalability: AI systems have almost no constr	ai	nts
	ai
	ai
	artificial intelligence	. Our response speed to simple
	AI	systems do not have to
if two	AI	systems are engaged in a
	ai
or analogy for reasoning about	AI	. This may lead
humans and	AI	to perform complex tasks. Resulting
	ai
for understanding the intelligence of	AI	systems. For us it is
	ai
	ai
	ai
	ai
So, if there would exist	AI	systems with general intelligence that
if we manage to construct	AI
human-	AI	teaming. Unless we decide to
capabilities of	AI	systems (which would not be
	ai
and that make the human-	AI	system stronger
Instead of pursuing human-level	AI	it would be more
	ai
	AI	has already established excellent capacities
efficient than biological intelligence.	AI	may thus help to produce
that ultimately the development of	AI	systems for supporting
people and	AI	systems will have to be
appeal to capacities in which	AI	systems excel, will have to
	ai
be required.	AI	systems are already much better
qualities	AI	systems may effectively take over
	ai
	ai
	ai
people are better suited than	AI	systems for a much broader
example, it is difficult for	AI	systems to interpret human
difficult to achieve within	AI	. As a result of all
	ai
with the overall limitations of	AI
limitations of humans and	AI	systems, human decisional biases may
	ai
collaboration between humans and	AI	that have developed the same
Although cooperation in teams with	AI	systems may need
	ai
most effective and safe human-	AI	systems (Elands et al., 2019
for general intelligence; instead of	ai
	ai
	ai
difficult to provide deep learning	AI
	ai
	ai
	ai
trust the results generated by	AI	. Like
technology, (e.g. Modeling & Simulation),	AI
	ai
trust in	AI	should be primarily based on
trickable) impressions, stories, or images	ai
	ai
The Impact of Multiple Narrow	AI	Technology
	ai
to narrow (limited, weak, specialized)	AI	. An AGI
	ai
	ai
characteristic of the current (narrow)	AI	tools is that they are
well-defined and structured. Narrow	AI	systems
circumstances, the adequacy of current	AI	is considerably reduced
Horowitz, 2018). As with narrow	AI
	ai
Multiple Narrow	AI	is Most Relevant Now
most crucial factor in future	AI	R&D, at least for
we tend to consider narrow	AI	applications as
of emerging	AI	applications will also have a
non-human-like)	AI	variants that will excel in
multiple variants of narrow	AI	applications also gradually get more
broader realm of integrated	AI	applications may be expected. In
to train a language model	AI
	ai
	AI	) can do this with
implies that the development of	AI
	ai
fruitful AI applications will m	ai	nly involve supplementing
	ai
narrow	AI	systems will be the major
driver of	AI	impact on our society
future, this may imply that	AI
	ai
concept. All dimensions of	AI	unfold and grow along their
of specific (narrow)	AI	capacities may gradually match, overtake
	AI	, for example in the field
	ai
So when	AI	will truly understand us as
of multiple forms of (integrated)	AI	systems. This concerns
	ai
performance and safety of human-	AI	systems (Peeters et al., 2020
being. According to most AI scientists, this will cert	ai	nly happen
system level, however, multiple narrow	AI
human and	artificial intelligence
	ai
	ai
the most probable potentials of	AI	applications for the
	ai
differences between natural and	artificial intelligence
very realistic and useful to	ai
most profitable	AI	applications for the short- and
be based on multiple narrow	AI	systems. These multiple
narrow	AI	applications may catch up with
AGI question, whether or when	AI	will outsmart us, take our
multiple	AI	innovations with humans as a
goals for	AI	systems (Elands et al., 2019
policy making the most fruitful	AI	applications
	ai
are safe to leave to	AI	and when is human judgment
and how to deploy	AI	systems effectively to complement and
	ai
No matter how intelligent autonomous	AI	agents become in
	ai
	ai
possible and specific variants of—	AI	systems. Only when humans develop
on the potential benefits of AI in (future) human-	AI	teams
relative to	AI	systems, the first challenge becomes
the more rigid abilities of	AI	?^4 In other words
and	artificial intelligence
	ai
	ai
characteristics, idiosyncrasies, and capacities of	AI	systems. This
possibilities, and limitations of the	AI	’s cognitive
	ai
and learning environments for human-	AI	systems. These flexible
	ai
	AI	systems and how to deal
performance, and choices of	AI	? Which may in some cases
the simple notion that	AI
are safe to leave to	AI	and when is human judgment
	AI	systems will grow
	ai
of	AI	. This is basically a subject
Cognitive Science of	AI	” may involve a range of
operation of the AI operating system or the “	AI
by AI,	AI	cognition (memory, information processing, problem
biases), dealing with	AI	possibilities and limitations in the
relation to cost-benefit), AI ethics and	AI	security. In
the working of the	AI	operating system. Due to the
which the	AI	technology and application develops, the
training-	ai
educational curricula on	AI	awareness. These subtopics go beyond
	AI	applications (i.e. conventional human
underlying system characteristics of the AI (the “	AI
	ai
specific qualities and limitations of	AI
the perspective of	AI	systems
biases in	AI
associated with the control of	AI
predictability of AI behavior (decisions), building trust, maint	ai	ning
with possibilities and limitations of	AI	in the field of
creativity”, adaptability of	AI	, “environmental awareness”, and
possible errors of	AI	which may be difficult to
Trust in the performance of	AI	(possibly in spite of limited
	ai
capitalize on the powers of	AI	in order to deal with
	ai
integrated combination of human- and	AI	faculties may perform at
enormous speed with which the	AI	technology
or deploy	AI	in relation to tasks and
	ai
and learning environments for human-	AI	systems are
	ai
changes in the field of	AI	and with the wide
1Narrow	AI	can be defined as the
	ai
3Unless of course AI will be deliberately constr	ai	ned or
the issue of Human-Aware AI, i.e. tuning	AI	to
Ackermann, N. (2018).	Artificial Intelligence	Framework: a visual
introduction to machine learning and	AI	Retrieved from
introduction-to-machine-learning-and-	ai	-d7e36b304f87. (September 9
Hybrid cognitive-affective Strategies for	AI
	ai
Bergstein, B. (2017).	AI	isn’t very smart yet
com/s/609318/the-great-	ai	-paradox
and Garrett, D. (2014). “Raising	AI
	ai
	ai
	ai
	ai
	ai
	ai
	ai
	ai
	ai
of	artificial intelligence	. Bulletin of the atomic scientists
	ai
	ai
	ai
	ai
	ai
	artificial intelligence	. Cham, Switzerland: Springer
	ai
intelligence in a human–AI society.	AI	and Society 38
E., and Knight, K. (1991).	Artificial intelligence	. 2nd edition
	ai
S., and Norvig, P. (2014).	Artificial intelligence	: a modern
	ai
B. (2020a). Design lessons from	AI	’s two grand goals
Shneiderman, B. (2020b). Human-centered	artificial intelligence
	ai
	ai
and Bronkhorst, K. (2018). Human-	AI	cooperation to
human-	AI	Co-learning. Adaptive instructional systems
Keywords: human intelligence,	artificial intelligence	, artificial
general intelligence, human-level	artificial intelligence	, cognitive
complexity, narrow artificial intelligence, human-	AI	collaboration
	ai	.2021.622364
	ai
	ai