|
3. Artificial Intelligence (AI)
171
their processes (adaptation and acceleration) and
thus in improving efficiency (in terms of costs and/or
personnel) or increasing flexibility as well as manag-
ing complexity and knowledge. The main objectives
with automation are to increase the percentage of
routine tasks that are automated and to bring about
a general improvement in system autonomy (e.g. au-
tonomous driving, firewalls). Within IT itself, soft-
ware automation (via learning) plays a key role. The
major areas of focus in process optimisation include
improving existing systems (adaptation), accelerat-
ing processes and thus saving time, and enhancing
quality (e.g. of forecasts). For the companies in-
volved, improving efficiency primarily means cutting
costs, but also increasing flexibility. Amongst other
things, they want to handle complexity more effec-
tively with the help of adaptive/learning systems
(e.g. security) and/or data science methods (dealing
with large volumes of data). Better knowledge man-
agement, i.e. gaining new insights from large data
volumes and spotting connections, is another import-
ant factor.
167
Innovations (new products and services) are a par-
ticularly strong motivation for Austrian companies to
use AI. A look at the applications that firms have de-
veloped to date reveals a broad picture. There is a
whole range of applications that cover speech and
language, dialogue systems (chatbots, assistance
systems, smart searching, etc.) or that analyse text
documents, manage knowledge or extract it (trend
and risk analysis for documents, data classification,
etc.).
There are also numerous applications connected
with industrial automation and process/plant engi-
neering (factory automation, Industry 4.0, system
optimisation, predictive maintenance, simulation in
production, engineering tools, analysis in production,
sensor fusion, etc.). Other applications are used to
classify and analyse image and video data (with ma-
ny centred around automation/autonomous opera-
167 See Prem and Ruhland (2019).
168 See Schaper-Rinkel (2019).
169 See Schaper-Rinkel (2019).
tion, especially autonomous driving) or optimise
transport/logistics (rolling stock optimisation, train
scheduling, etc.). IT itself is another area of applica-
tion for AI technology, e.g. in the fields of soft-
ware-defined networks, software management, secu-
rity (IT systems) and making sensitive personal data
anonymous. Finally, AI at Austrian companies can al-
so be found in risk management, controlling and, in
many cases, data analysis. The AI technologies used
here mainly comprise machine learning, data analysis
and forecasting techniques, speech processing, im-
age analysis, and deductive and knowledge-based
systems.
AI can have an innovative effect in various ways. It
is seen as having great economic potential (produc-
tivity and price impact), particularly with regard to
the automation of routine activities, while also being
capable of forming the basis for enhanced and/or
new products and services. Companies can harness
the potential offered by AI in various ways. Knowl-
edge can either be developed chiefly in house or
bought in from outside, And there would appear to
be many different possible gradations between these
two extremes.
168
Being both so popular and so disruptive, AI will
offer a great deal of potential and bring a great deal
of impact – neither of which will be particularly easy
to forecast – for a large number of industries and
companies. Besides its ramifications within a compa-
ny itself, AI will also cause shifts within and between
industries and thus drive forward structural change.
Companies often view AI as a sub-field and combine
it with other digitalisation issues and strategies,
which causes boundaries to be blurred. This is com-
pounded by legislative and regulatory grey areas,
which can either accelerate or curb the use of AI. The
main technical challenges relate to access, availabil-
ity and quality as well as the processing of data in AI
systems, system architectures and aspects of securi-
ty, data protection and privacy (e.g. personal data).
169
Do'stlaringiz bilan baham: |
