Machine learning and artificial intelligence – pure hype (?) - part 1
Artificial intelligence and machine learning – no matter the medium you use (print media, online portals, radio or TV), you will generally come across one of these terms sooner or later. But as the saying goes – everything has already been said, but not yet by everybody.
In this first blog post of our series I will examine our views on the topic as data analysts, and give you a sense of how you can use these methods (and our solutions) in this connection in order to generate real added value for you and/or your company.
The focus here is on practical application, rather than conveying specialist technical principles. Given the extensive scope of the topic, I will divide the blog into several sections:
- Introduction: From global galactic issues to audit, ICS, data analytics and risk management (focus in this blogpsot)
- Building bridges: why initial disappointment may set in – from your data to machine learning algorithms
- Content is king (here too): Creativity comes later – only actual use will create added value.
- Three examples of use: a – Old wine in new bottles; b – A new approach; c – Doing things differently
- Summary
Before going into details, I should first like to examine the terms artificial intelligence and machine learning:
- Artificial intelligence: A sub-area of informatics, dealing with the automation of intelligent behaviour.
- Machine learning: A sub-area of artificial intelligence that aims to generate artificial knowledge from experience, and make this in itself usable for the above automation of behaviour.
1. Introduction
Like me, you probably think there's no escaping talk of AI/ML, even if you're not involved with data analytics, audit, internal control system or risk management in your professional life.
“Artificial intelligence” and “machine learning” are clearly very prominent topics right now. One evening, when flicking through conventional TV programmes, Bavarian Minister President Markus Söder appears on screen, (rightly) stating that we mustn't find ourselves lagging behind our international competitors – and highlighting, among other things, the importance of flagship projects in the field of artificial intelligence. The next morning in the car on the way to work, a report says that artificial intelligence is here to stay. Yet company decision-makers will still always need emotions and professional experience (or a gut feeling) to correctly interpret and accept or reject proposals made by AI. Then at work, while settling myself at my desk with a cup of coffee, my web browser shows me the headline, on Spiegel.de, “Little demand for AI skills at DAX-listed companies”. As I reach for my mobile phone to open my favourite apps, Google Photos shows me a photo collage entitled “How quickly they grow up”, depicting my eldest son’s first 4 years. After closing this notification, Siri on my home screen tries to suggest which call I might like to make next (birthday wishes for a friend). Getting up out of my chair and moving across the office to look contemplatively out of the window, I cast a glance at Deggendorf Institute of Technology on the opposite side of the street. At the start of October, it successfully launched the new course “Artificial Intelligence”. And I won’t even mention the embarrassingly funny suggestions given to me by Amazon based on a combination of my browser search history, my Prime watchlist, and previous purchases. I just want to say “I can explain that.” I was actually searching for clips with weights to secure garden furniture covers at home against the wind. The resulting suggested purchase is quite another story..
It's a fact that current discussions on these topics tend to be very general. It's also clear that we are specifically confronted with them in our daily use of apps, computers, mobile phones and vehicles.
However, our readers and customers (and therefore presumably very likely you) are specifically interested in matters such as audit, accounting, internal control system, risk management, GRC and data analytics. What relevance does machine learning/AI have here?
This blog will deal with the “here and now” in 2020. I don't want to rule out the possibility that in 2, 3, 5 or 10 years’ time AI-based solutions may have completely automated many of the tasks in these areas and be able to perform them largely without any human input at all (I don't have my crystal ball on me just now). Personally, I believe it's important that as well as pondering possible future developments, it’s also important to consider the existing, currently utilisable potential.
I hope you enjoyed the first part of our blog series. In the next post we will take a closer look at the machine learning algorithms.