New ways of working
Week 1
Digital economy by Orlikowski and C. Suzanne Iacono, 2000
Aim of the article: analyse the digital economy from a microsocial and organisational
perspective. The authors argue that digital economy is a social product enacted by
intertwining new technologies within practises and processes
Social predictions encourage new ways of thinking but such predictions are problematic
because:
- Mislead on a factual level by generalising the cause. Generalisations are never
accurate
- Mislead on a theoretical level. It points that digital economy creates independency
between technology and social forces. Change is however made of complex
connections
Fallacies in theories:
- Technological determinism: technology is view as an external entity that determines
or forces change within the social system. By measuring and modelling the changes
caused by technology, future changes can be predicted.
o The same technology has different outcomes based on different
circumstances
- Strategic choice: technology is a malleable source that can put into a variety of uses.
We can predict changes by focusing on identifying motivations and objective intrinsic
to technology. The choice of the product determines the outcome
o Users shape the use of technology on their own need by developing
“workarounds”
Proposed theory solution Enacted approach
See the relations between technology and organisations as an ongoing phenomena where
actors actively influence each other.
“Rather, the organizational changes associated with the use of technologies are
shaped by human actions and choices, while at the same time having consequences
that we cannot fully anticipate or plan”
Digital economy is a ongoing product shaped and produced by humans and organisations
which both have intended and unintended consequences on each other by equally shaping
their actions.
,Implications of using technologies in organisations
Social dynamic and multiple
Technology is a product and a medium of human action
- Technologies are social because they are constructed by people
- Technologies are dynamic and not stable but provisional
- Technology is multiple because is formed by a variety of tools and configurations
that are interlinked with each other
The effect is varied, embedded and emergent
- Failing to pay attention on what people actually do with a technology leads to focus
on the wrong technology characteristics (artifact itself, features, discourse etc..)
- Organisations tend to make assumptions about technology use ignoring the “right
use” of it. This is because technology has 2 dimensions embedded
o Espoused technology: expectations about the functions and features within
the technology
o Technologies-in-use: the ways we actually use specific technological
affordances based on our skills, tasks and purposes (that vary) day by day.
Focuses on work practices. Thus, we cannot predict the actual use of a
technology by looking at the espoused technology dimension
this distinction addresses the debate around the productivity paradox:
the idea that the increased investment in IT is not producing increased
productivity. IT cannot increase productivity per se, but only the use of
technology can.
- Technology is emergent because we constantly make choices about
whether/how/why use technology. If technology does not benefit us we abandon it,
change it, invent new tools
Unintended consequences
Our actions have unintended consequences and multiple implications on what is around us,
Digital economy must take into account the consequences of living and working. It is
separate from intentions during the design, use and immediate context. performativity
Organisations and the digital economy
Reason why organisations are engaging with digital economy
First answer (more common)
- Resource needs of organisations
- Expectations of reduced costs
- Access to new markets
- Cost effective internetworking technologies
,Second answer
- Strategic choice: both in terms of economic analysis of information flow and
resource dependence between organisations
o Organisations need to be flexible because a 360 internal learning is not
possible anymore aka they need to partner and collaborate with others
- Technological determinism: causal relationship between technological infrastructure
and the social architecture/economy/organisation
o From industrial activity and modernist systems to information and
postmodern systems
Issues with these 2 approaches
They ignore the difficulties in implementing technological change and challenges of
dealing with consequences
Internetworking, for example, raises issues oh how permeable the organisation boundaries
should be. Organisational engagement consists of:
- Communicating via email
- Generating a web presence
- Establishing buyer-supplier transaction networks
- Create real time virtual integration (the simple presence of technology does not
guarantee success)
We also do not know which organizations are not connecting and why, and what
types of challenges organizations face when attempting to participate in the Internet. What
it means for organizations to “be on the Internet” will also evolve as new technologies,
business models, regulations, laws, and organizational processes emerge
Conclusion
The enacted view suggests that to assess the future of the organisations we have to
consider:
- Time, context, and technology-specific generalisations
- Choose how and why we use internetworking technologies shape and
consequences of the digital economy
- Answer the question: what sort of digital economy do we want to create?
When the machine meets the expert 2022
Aim of the paper: How ML changes our understanding of knowledge production in
organisations and how ML developers manage the balance between independence and
relevance?
hybrid practice of mutual learning: developers and domain experts reflect and
adapt the activities
Machine learning (ML): a broad set of techniques of AI that can derive and revise knowledge
automatically over time by learning from data
, Current ML systems attempt to automate knowledge work by learning from data without
relying on the involvement of domain experts since those insights are superior quality
speaking.
different from traditional systems because they
- are independent
- mitigate human biases, inefficiencies and path dependences
Supervised learning: the algorithm learns from a labelled training dataset to predict the
labels for future data
Unsupervised learning: there are no labels to predict and instead the algorithm groups or
clusters the data in some meaningful way
Issue within the ML
novel tension between independence and relevance
Independence: producing knowledge without relying on domain expertise
Relevance: producing knowledge that is useful to the domain
ML paradigm
Knowledge can be produced through a set of activities different from traditional IT systems
to realise new promises for organisations. These activities represent a radical empiricist
mode of knowledge production that attempts to avoid the ned for expert work in favor of
independence.
HOWEVER
This pursuit of independence is problematic as ML systems fail to recognise what is
relevant to a particular course of action and a broader domain
IT activities ML activities ML promises
Elicit knowledge from experts Collect a large set of training Objective knowledge free
through communication data with labelled examples from biases and subjectivity
techniques
Codify knowledge from Train a learning algorithm on Produce new knowledge that
experts into explicit the data to derive knowledge expertise might not conceive
procedures, such as rules and in the form of a predictive or anticipate
constraints model
Share and apply newly Deploy the model to predict Produce knowledge with an
produced knowledge future outcomes efficiency and speed that
exceeds experts abilities
Issues with ML promises:
- frame problem: inability of algorithms to reason about and act on events thet are
not design to handle
- account for the broader social system that surrounds applications
IT activity failure:
to deliver their promises.
This problem occurs because of misalignment between the idea of what constitutes knowledge and
how knowledge actually manifests in experts daily work.
Week 1
Digital economy by Orlikowski and C. Suzanne Iacono, 2000
Aim of the article: analyse the digital economy from a microsocial and organisational
perspective. The authors argue that digital economy is a social product enacted by
intertwining new technologies within practises and processes
Social predictions encourage new ways of thinking but such predictions are problematic
because:
- Mislead on a factual level by generalising the cause. Generalisations are never
accurate
- Mislead on a theoretical level. It points that digital economy creates independency
between technology and social forces. Change is however made of complex
connections
Fallacies in theories:
- Technological determinism: technology is view as an external entity that determines
or forces change within the social system. By measuring and modelling the changes
caused by technology, future changes can be predicted.
o The same technology has different outcomes based on different
circumstances
- Strategic choice: technology is a malleable source that can put into a variety of uses.
We can predict changes by focusing on identifying motivations and objective intrinsic
to technology. The choice of the product determines the outcome
o Users shape the use of technology on their own need by developing
“workarounds”
Proposed theory solution Enacted approach
See the relations between technology and organisations as an ongoing phenomena where
actors actively influence each other.
“Rather, the organizational changes associated with the use of technologies are
shaped by human actions and choices, while at the same time having consequences
that we cannot fully anticipate or plan”
Digital economy is a ongoing product shaped and produced by humans and organisations
which both have intended and unintended consequences on each other by equally shaping
their actions.
,Implications of using technologies in organisations
Social dynamic and multiple
Technology is a product and a medium of human action
- Technologies are social because they are constructed by people
- Technologies are dynamic and not stable but provisional
- Technology is multiple because is formed by a variety of tools and configurations
that are interlinked with each other
The effect is varied, embedded and emergent
- Failing to pay attention on what people actually do with a technology leads to focus
on the wrong technology characteristics (artifact itself, features, discourse etc..)
- Organisations tend to make assumptions about technology use ignoring the “right
use” of it. This is because technology has 2 dimensions embedded
o Espoused technology: expectations about the functions and features within
the technology
o Technologies-in-use: the ways we actually use specific technological
affordances based on our skills, tasks and purposes (that vary) day by day.
Focuses on work practices. Thus, we cannot predict the actual use of a
technology by looking at the espoused technology dimension
this distinction addresses the debate around the productivity paradox:
the idea that the increased investment in IT is not producing increased
productivity. IT cannot increase productivity per se, but only the use of
technology can.
- Technology is emergent because we constantly make choices about
whether/how/why use technology. If technology does not benefit us we abandon it,
change it, invent new tools
Unintended consequences
Our actions have unintended consequences and multiple implications on what is around us,
Digital economy must take into account the consequences of living and working. It is
separate from intentions during the design, use and immediate context. performativity
Organisations and the digital economy
Reason why organisations are engaging with digital economy
First answer (more common)
- Resource needs of organisations
- Expectations of reduced costs
- Access to new markets
- Cost effective internetworking technologies
,Second answer
- Strategic choice: both in terms of economic analysis of information flow and
resource dependence between organisations
o Organisations need to be flexible because a 360 internal learning is not
possible anymore aka they need to partner and collaborate with others
- Technological determinism: causal relationship between technological infrastructure
and the social architecture/economy/organisation
o From industrial activity and modernist systems to information and
postmodern systems
Issues with these 2 approaches
They ignore the difficulties in implementing technological change and challenges of
dealing with consequences
Internetworking, for example, raises issues oh how permeable the organisation boundaries
should be. Organisational engagement consists of:
- Communicating via email
- Generating a web presence
- Establishing buyer-supplier transaction networks
- Create real time virtual integration (the simple presence of technology does not
guarantee success)
We also do not know which organizations are not connecting and why, and what
types of challenges organizations face when attempting to participate in the Internet. What
it means for organizations to “be on the Internet” will also evolve as new technologies,
business models, regulations, laws, and organizational processes emerge
Conclusion
The enacted view suggests that to assess the future of the organisations we have to
consider:
- Time, context, and technology-specific generalisations
- Choose how and why we use internetworking technologies shape and
consequences of the digital economy
- Answer the question: what sort of digital economy do we want to create?
When the machine meets the expert 2022
Aim of the paper: How ML changes our understanding of knowledge production in
organisations and how ML developers manage the balance between independence and
relevance?
hybrid practice of mutual learning: developers and domain experts reflect and
adapt the activities
Machine learning (ML): a broad set of techniques of AI that can derive and revise knowledge
automatically over time by learning from data
, Current ML systems attempt to automate knowledge work by learning from data without
relying on the involvement of domain experts since those insights are superior quality
speaking.
different from traditional systems because they
- are independent
- mitigate human biases, inefficiencies and path dependences
Supervised learning: the algorithm learns from a labelled training dataset to predict the
labels for future data
Unsupervised learning: there are no labels to predict and instead the algorithm groups or
clusters the data in some meaningful way
Issue within the ML
novel tension between independence and relevance
Independence: producing knowledge without relying on domain expertise
Relevance: producing knowledge that is useful to the domain
ML paradigm
Knowledge can be produced through a set of activities different from traditional IT systems
to realise new promises for organisations. These activities represent a radical empiricist
mode of knowledge production that attempts to avoid the ned for expert work in favor of
independence.
HOWEVER
This pursuit of independence is problematic as ML systems fail to recognise what is
relevant to a particular course of action and a broader domain
IT activities ML activities ML promises
Elicit knowledge from experts Collect a large set of training Objective knowledge free
through communication data with labelled examples from biases and subjectivity
techniques
Codify knowledge from Train a learning algorithm on Produce new knowledge that
experts into explicit the data to derive knowledge expertise might not conceive
procedures, such as rules and in the form of a predictive or anticipate
constraints model
Share and apply newly Deploy the model to predict Produce knowledge with an
produced knowledge future outcomes efficiency and speed that
exceeds experts abilities
Issues with ML promises:
- frame problem: inability of algorithms to reason about and act on events thet are
not design to handle
- account for the broader social system that surrounds applications
IT activity failure:
to deliver their promises.
This problem occurs because of misalignment between the idea of what constitutes knowledge and
how knowledge actually manifests in experts daily work.
