36
ways not envisaged by traditional cloud computing. Similarly, cloud computing is becoming more heterogeneous and composable with a wider
variety of customisable configurations available to clients that impact performance and complicate service level expectations, thus pushing
performance- related decisions to the client, and requiring more nuanced
agreements. Furthermore, with the advent of the Internet of Things, the
cloud is becoming more decentralised and distributed across a cloud-tothing (C2T) continuum. This has resulted in new computing paradigms
including fog, mist and edge computing (Iorga et al. 2018). Processing
and storage may take place in the cloud, at the edge or somewhere in
between (the fog).
This new decentralised, abstract, heterogeneous, and composable cloud
introduces complexity at several orders of magnitude higher than today. It
is beyond human capabilities to manage such infrastructure manually. As a
result, the cloud is becoming even more automated and intelligent.
Artificial Intelligence for IT Operations (AIOps) algorithms and machine
learning monitor, operate, and maintain distributed systems (Cordoso
2019). The emergence of self-organising and self-learning systems represents a significant evolution in cloud infrastructure decision-making.
Outsourcing decision-making to AI provides substantial technical, legal
and trust challenges, not least the black box nature of AI decision making.
It is foreseeable that the actions of AI will result in cloud under- performance
at some time in the future and addressed in accordance with existing legal
provisions. However commentators have noted that AI may not be recognised as a subject of law, and as a result, may not be held personally liable
for the damage it causes (C ̌ erka et al. 2015). Cloud contracts need to
evolve to reflect this changing and more nuanced cloud.
At the same time, the nature of contracts is developing, albeit at a much
slower pace. Spulber (2018) has proposed a new framework for ‘intellectual contracts’, a form of “…agreement to create, develop, share, or apply
intangible assets involved in technological change.” In his conceptualisation, Spulber attempts to overcome the shortcomings of traditional contracts with respect to the completeness, excludability, and transferability of
intangible assets while recognising that IP arises from intentional and
unintentional cooperation, and rights in such outputs needs to be
addressed in contracts. Similarly, there has been renewed discussions on
the value of smart contracts in cloud computing with the emergence and
hype around Blockchain. Smart contracts are not new; in effect they are
agreements whose execution is automated and self-enforceable. Vending
T. LYNN
ways not envisaged by traditional cloud computing. Similarly, cloud computing is becoming more heterogeneous and composable with a wider
variety of customisable configurations available to clients that impact performance and complicate service level expectations, thus pushing
performance- related decisions to the client, and requiring more nuanced
agreements. Furthermore, with the advent of the Internet of Things, the
cloud is becoming more decentralised and distributed across a cloud-tothing (C2T) continuum. This has resulted in new computing paradigms
including fog, mist and edge computing (Iorga et al. 2018). Processing
and storage may take place in the cloud, at the edge or somewhere in
between (the fog).
This new decentralised, abstract, heterogeneous, and composable cloud
introduces complexity at several orders of magnitude higher than today. It
is beyond human capabilities to manage such infrastructure manually. As a
result, the cloud is becoming even more automated and intelligent.
Artificial Intelligence for IT Operations (AIOps) algorithms and machine
learning monitor, operate, and maintain distributed systems (Cordoso
2019). The emergence of self-organising and self-learning systems represents a significant evolution in cloud infrastructure decision-making.
Outsourcing decision-making to AI provides substantial technical, legal
and trust challenges, not least the black box nature of AI decision making.
It is foreseeable that the actions of AI will result in cloud under- performance
at some time in the future and addressed in accordance with existing legal
provisions. However commentators have noted that AI may not be recognised as a subject of law, and as a result, may not be held personally liable
for the damage it causes (C ̌ erka et al. 2015). Cloud contracts need to
evolve to reflect this changing and more nuanced cloud.
At the same time, the nature of contracts is developing, albeit at a much
slower pace. Spulber (2018) has proposed a new framework for ‘intellectual contracts’, a form of “…agreement to create, develop, share, or apply
intangible assets involved in technological change.” In his conceptualisation, Spulber attempts to overcome the shortcomings of traditional contracts with respect to the completeness, excludability, and transferability of
intangible assets while recognising that IP arises from intentional and
unintentional cooperation, and rights in such outputs needs to be
addressed in contracts. Similarly, there has been renewed discussions on
the value of smart contracts in cloud computing with the emergence and
hype around Blockchain. Smart contracts are not new; in effect they are
agreements whose execution is automated and self-enforceable. Vending
T. LYNN