{"id":1110,"date":"2020-07-10T19:22:59","date_gmt":"2020-07-10T19:22:59","guid":{"rendered":"http:\/\/janbosch.com\/blog\/?p=1110"},"modified":"2020-07-10T19:23:01","modified_gmt":"2020-07-10T19:23:01","slug":"the-ai-of-digitalization","status":"publish","type":"post","link":"https:\/\/janbosch.com\/blog\/index.php\/2020\/07\/10\/the-ai-of-digitalization\/","title":{"rendered":"The AI of digitalization"},"content":{"rendered":"\n
\"\"
Image by Pete Linforth from Pixabay<\/figcaption><\/figure>\n\n\n\n

This article is the last of four where I explore different dimensions of digital transformation. Earlier, I discussed business models<\/a>, product upgrades<\/a> and data exploitation<\/a>. The fourth dimension is concerned with artificial intelligence. Similar to the other dimensions, our research showed that there\u2019s a clear evolution path that companies go through as they transition from being traditional companies to becoming digital ones (see the figure).<\/p>\n\n\n\n

In the first stage, the company is still focused on data analytics. \nAll data is processed for the sole purpose of human consumption and \ninterpretation. At this point, things are all about dashboard, \nvisualization and stakeholder views.<\/p>\n\n\n\n

\"\"\/
Evolution of the use of AI technology.<\/figcaption><\/figure>\n\n\n\n

In the second stage, the first machine learning (ML) or deep learning\n (DL) models are starting to be developed and deployed. The training of \nthe models is based on static data sets that have been assembled at one \npoint in time and that don\u2019t evolve unless there\u2019s an explicit decision \ntaken. When that happens, a new data set is assembled and used for \ntraining.<\/p>\n\n\n\n

In the third stage, DevOps and MLOps are merged in the sense that \nthere\u2019s a continuous retraining of models based on the most recent data.\n This data is no longer a data set, but rather a window over a data \nstream that\u2019s used for training and continuous re-training. Depending on\n the domain and the rate of change in the underlying data, the MLOps \nloop is either aligned with the DevOps loop or is executed more or less \nfrequently. For instance, when using ML\/DL for house price prediction in\n a real-estate market, it\u2019s important to frequently retrain the model \nbased on the most recent sales data as house prices change continuously.<\/p>\n\n\n\n

Especially in the software-intensive embedded systems industry, as \nML\/DL models are deployed in each product instance, the next step tends \nto be the adoption of federated approaches. Rather than conducting all \ntraining centrally, the company adopts federated learning approaches \nwhere all product instances are involved in training and model updates \nare shared between product instances. This allows for localization and \ncustomization as specific regions and users may want the system to \nbehave differently. Depending on the approach to federated learning, \nit\u2019s feasible to allow for this. For example, different drivers want \ntheir adaptive cruise control system to behave in different ways. Some \nwant to have the system take a more careful approach whereas others \nwould like to see a more aggressive way of breaking and accelerating. \nEach product instance can, over time, adjust itself in response to \ndriver feedback.<\/p>\n\n\n\n

Finally, we reach the automated experimentation stage where the \nsystem fully autonomously experiments with its own behavior with the \nintent of improving certain success metrics. Whereas in earlier stages, \nhumans conduct A\/B experiments or similar and the humans are the ones \ncoming up with the A and B alternatives, here it\u2019s the system itself \nthat generates alternatives, deploys, measures the effect and decides on\n next steps. Although the examples in this category are few and far \nbetween, we\u2019ve been involved in, among others, cases where we use a \nsystem of this type to explore configuration parameter settings (most \nsystems have thousands) in order to optimize the system\u2019s performance \nautomatically.<\/p>\n\n\n\n

Concluding, digital transformation is a complex, multi-dimensional challenge. One of the dimensions is the adoption of AI\/ML\/DL. Using AI is not a binary step, but rather a process that evolves over time and proceeds through predefined steps. Deploying AI allows for automation of tasks that couldn\u2019t be automated earlier and for improving the outcomes of automated processes through smart, automated decisions. Once you have software, you can generate data. Once you have data, you can employ AI. Once you have AI, you can truly capitalize on the potential of digitalization.<\/p>\n\n\n\n

This blog series is taking a summer break and will be back in mid August.<\/em><\/p>\n\n\n\n

To get more insights earlier, sign up for my newsletter at\u00a0<\/em>jan@janbosch.com<\/em><\/a> or follow me on<\/em> janbosch.com\/blog<\/em><\/a>, LinkedIn (<\/em>linkedin.com\/in\/janbosch<\/em><\/a>) or Twitter (<\/em>@JanBosch<\/em><\/a>).<\/em><\/p>\n","protected":false},"excerpt":{"rendered":"

This article is the last of four where I explore different dimensions of digital transformation. Earlier, I discussed business models, product upgrades and data exploitation. The fourth dimension is concerned with artificial intelligence. Similar to the other dimensions, our research showed that there\u2019s a clear evolution path that companies go through as they transition from … Read more<\/a><\/p>\n","protected":false},"author":2,"featured_media":0,"comment_status":"open","ping_status":"open","sticky":false,"template":"","format":"standard","meta":{"generate_page_header":"","footnotes":""},"categories":[15,10],"tags":[],"_links":{"self":[{"href":"https:\/\/janbosch.com\/blog\/index.php\/wp-json\/wp\/v2\/posts\/1110"}],"collection":[{"href":"https:\/\/janbosch.com\/blog\/index.php\/wp-json\/wp\/v2\/posts"}],"about":[{"href":"https:\/\/janbosch.com\/blog\/index.php\/wp-json\/wp\/v2\/types\/post"}],"author":[{"embeddable":true,"href":"https:\/\/janbosch.com\/blog\/index.php\/wp-json\/wp\/v2\/users\/2"}],"replies":[{"embeddable":true,"href":"https:\/\/janbosch.com\/blog\/index.php\/wp-json\/wp\/v2\/comments?post=1110"}],"version-history":[{"count":1,"href":"https:\/\/janbosch.com\/blog\/index.php\/wp-json\/wp\/v2\/posts\/1110\/revisions"}],"predecessor-version":[{"id":1112,"href":"https:\/\/janbosch.com\/blog\/index.php\/wp-json\/wp\/v2\/posts\/1110\/revisions\/1112"}],"wp:attachment":[{"href":"https:\/\/janbosch.com\/blog\/index.php\/wp-json\/wp\/v2\/media?parent=1110"}],"wp:term":[{"taxonomy":"category","embeddable":true,"href":"https:\/\/janbosch.com\/blog\/index.php\/wp-json\/wp\/v2\/categories?post=1110"},{"taxonomy":"post_tag","embeddable":true,"href":"https:\/\/janbosch.com\/blog\/index.php\/wp-json\/wp\/v2\/tags?post=1110"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}