Finance, technology and disruption

Abstract

In this paper, we assess how recent technology advances have changed the way we coordinate. After a brief discussion of the common challenges to effective coordination, we highlight some important implications of technology on addressing informational and behavioral frictions. We focus on discussing the effects of three specific technology developments including artificial intelligence (AI), automation, and blockchain, on the choice of coordination modes. We argue that technology is shifting the boundaries between firms and markets and is opening the door to new research directions.

KEYWORDS:

• Coordination
• technology
• boundaries of firm
• AI
• automation
• blockchain

Disclosure statement

No potential conflict of interest was reported by the authors.

ORCID

Raghavendra Rau http://orcid.org/0000-0002-3320-5104

Notes

1 We limit our discussion to these two common forms of coordination and aim to provide general implications of the change in technology on coordinating economic activities. Other important coordination mechanisms, including vertical integration (Williamson 1971) and coordination through various forms of committees (Farrell and Saloner 1988), among others, are outside of the scope of this discussion.

2 Henceforth, we use coordination through hierarchies and firms interchangeably.

3 Even within a hierarchy, not all resource allocations are coordinated through command. For example, in Lancashire cotton industry, looms and yarn can be obtained on credit (Coase 1937).

4 Henceforth, we follow previous literature and use coordination through markets and the price mechanism interchangeably (Coase 1937; Williamson 1973).

5 The artificial intelligence literature contains many definitions of data ontology; many of these are contradictory (Noy and McGuinness 2001). For the purposes of this paper, we define data ontology as a formal explicit description of the characteristics (features and attributes) of observations in a dataset. We refer efficiency of data ontology here as the speed and accuracy of describing a feature or attribute of one or many observations in a given dataset, and the extent to which the end users can identify the observation with the corresponding data ontology at ease.

6 For more information on the five factor personality traits, see Costa and McCrae (1992).

7 Evidence so far shows that human behavior is predictable in isolation. However, much of human behavior also depends on the immediate environment, causing the formation of emergent behaviors. Evidence on whether emergent behaviors are predictable is currently limited.

8 See Christie (2018) ‘Artificial intelligence: Winter is coming’ in Financial Times.

9 A blockchain contains data and information in blocks with the same size. Each block stores the hashed data from the previous block, and new data generated in the current block-forming window, to provide cryptographic security. The hashing process uses SHA256 function, which is a one way hash process. It can transform most, if not all, types of data (perfect or not) into a hash key, e.g., a series of numbers. For example, written languages, transcripts of audio, relational data, descriptions of images (arguably the most common forms of unstructured data), can be transformed using the hash function and stored on a blockchain.

10 This discussion is adapted from Buterin (2014).

11 For discussions on blockchain and corporate governance, see Yermack (2017).