Digitization is already here. By today’s standards, every legal professional possesses his personal computer and smartphone and utilizes standard software. But is this already a digital transformation? Digital transformation, i.e. digitalization, does not only mean converting analog into digital information, but rather the integration of digital technology into all areas of business in the form of a fundamental transformation of how businesses operate. Therefore, it has the potential to include a positive paradigm shift towards a smart and technological world that emphasizes the creative input of human work. The resulting gains in efficiency were, however, accompanied by a change in expectations of employers, clients and business partners. The workload of (legal) professionals has increased even further because the possibility of automation raises expectations and the need for automation. The hope that digitization would enable humans to work differently with a shift from fast-paced mundane to creative and quality work has so far been proven wrong. Will a true digitalization in the sense of a paradigm shift and digital transformation bring us closer to turning that vision into reality?
Digitization has led to an abundance of information. This poses the difficult problem of information overload. The distinction between relevant and irrelevant data has become one of the main challenges of our informational age. Artificial Intelligence (AI) offers the possibility to cause a paradigm shift, freeing humans from simple and repetitive tasks by automation and informational analysis.
Automation is already in widespread use. For example, legal documents can be created automatically, eliminating common human mistakes such as misjudgments, omissions, missing references or even simple typos. Or document-intake including understanding its content and automated triage can be carried out by bots without much human contribution. This enables humans to channel their energy into more creative and complex tasks like designing legal processes or complex negotiations.
Informational analysis, however, is still a bottleneck. AI can already process millions of datasets. However, humans have a hard time using the information an AI system processes efficiently as they often lack a clear understanding of the results of the analysis and how it was calculated. AI and human intelligence “perceive” our reality differently. AI “thinks” in correlations, whereas humans mainly use causality to link data (Leetaru, A Reminder That Machine Learning Is About Correlations Not Causation, https://www.forbes.com/sites/kalevleetaru/2019/01/15/a-reminder-that-machine-learning-is-about-correlations-not-causation/?sh=535c6fc96161). We can overcome this hurdle by using visualization of the results and well-designed processes. Visualization can make it easier for humans to use the information provided by AI-processing and thus bridge the gap between the difference in AI’s and human’s perception of data.
In summary, digitalization, in particular technologies such as AI, have the potential to bring a paradigm shift upon human work. Humans may be relieved from repetitive work, allowing them to be more creative, and therefore more humane. What will be most important is to be innovative instead of productive.
Anzeige
Bitte loggen Sie sich ein, um Zugang zu Ihrer Lizenz zu erhalten.
Semmler and Rose (2017), pp. 85, 86 et seq. Also, the area of credit scoring is widely automated nowadays, compare Burell (2016), p. 2; Binns (2018), pp. 31, 543, 545.
Smaller firms can suddenly compete with big law firms utilizing technology instead of manpower, compare Semmler and Rose (2017), pp. 85, 90. Also, by using such innovative means new law firms can quickly establish a competitive market position, which is illustrated by the Big Four’s emergence as market competitors.
For instance, in the field of IT-security it is much more important to identify statistical outliers like dubious lump sum payments to single bank accounts, that are not usually payees. In contrast to that, predictive analytics usually focus on certain reoccurring patterns within person’s decisions to identify a certain modus operandi that can be relied upon as a general rule. Cf. Ahmed and Najmul Islam (2020), pp. 427, 430.
This might be done by the computer system automatically, which does not stipulate an act of original artificial intelligence though, because it will most likely be implemented as a guideline to the computer system by a human programmer. This is then an element of the in comparison to machine learning much older form of symbolic artificial intelligence, for it represents a relation that can easily be articulated with symbolic language.
In video-game AI’s it might for instance cause a trained AI-model to fail, just because a graphical element in the game has changed after the training process, Garnelo and Shanahan (2019), p. 17.
Sometimes called “dirty data”, Won Kim et al. (2003), p. 81. Even official data, like police data, can often not be relied upon, compare Richardson et al. (2019), pp. 192, 197 et seq.
For that reason AI-systems based on machine learning (especially deep neural networks) are often labeled as a blackbox in today’s AI-discourse, cf. Wischmeyer (2020), p. 75, passim; Garnelo and Shanahan (2019), pp. 29, 17; Ferguson (2017), pp. 1109, 1165.
Therefore, in the field of design and marketing, AI-technologies might be utilized, but never replace human intelligence altogether, see Liikkanen (2019), pp. 600–604, 603.
Cf. on that matter the interesting discussion of computing as a diagnostic and formalizing measure in effectuating social changes, Abebe et al. (2020), pp. 252–260, 253 et seq.
Kirste (2019), pp. 58, 67. As studies show this is also an important factor in the effectiveness of AI-suggestions, see inter alia Grgić-Hlača et al. (2019), pp. 1, 10 et seq.