Translation of the Original Article by Siegfried Lautenbacher 07.03.2016
Disclaimer: This article was created for the #Blogparade #schlauerarbeiten by Stefan Pfeiffer. The author has his own machine learning and natural language processing technologies with his Think Tank valuescope and his team at Beck et al. Services is currently working on Watson-based Apps. This Post is theoretical, critical “in a loving sense” and mercilessly subjective.
Knowledge is Power
This now old-fashioned sounding and almost scorned proverb is the shortened version of a sentence by Francis Bacon, which in its original version ("nam et ipsa scientia potestas est" from the meditationes sacrae 1597 German: because science itself is power) had a profoundly enlightening function. After all, this sentence contains the program of modernity, the starting point of enlightenment: Scientific knowledge, it was hoped, replaces at last and final the traditional instances of the interpretation of the world, such as religion with their priests and the rulers with their absolutist apparatus. We have really come far with this.
„Knowledge are commodities"
This is today’s valid economized version of the Baconian sentence and knowledge is treated as such: With industrialization after the Tayloristic model, the idea arose of being able to
- manage knowledge and to (re) produce it automated as possible and according to identical conditions
- to trade it , to buy or sell it
- to control it
- and to simply dispose of irrelevant knowledge.
This notion, incidentally, dominates our debate on education in my view. The apologists of the competence ideology point out that today it is no longer necessary to know something, but that the key to success lies in the application and utilization of the knowledge. Therefore, one could do without education at school in good cheer; the main thing for our children to learn is to learn. Lifelong learning becomes a mantra, a necessity, a compulsion, only no one knows any longer what actually is to be learned. Uncoupled and forgotten is the necessary condition for knowledge: the autonomous entity, which, as an individual in a social educational process (social learning), enables his capacity to act through reflection of knowledge and thereby changing it.
What does this have to do with Watson?
In elaborate advertising Watson presents itself currently human, all too human: "He / She / It" chats away with Bob Dylan or Serena Williams, visits Self-Help-groups, appears funny, informed, eloquent and self-deprecating. "Watson thinks with us" is the message at the end of the Spots, "to outthink the limits". With the focus on Watson, IBM touts a new era of computing for itself. It is about nothing less than about "Cognitive Business". The "Cognitive" I understand as an applied artificial intelligence, when it comes to deal with problems where the solution can be accomplished only with the help of intelligence, rather than by following of pre-established rules.
IBM has indeed a long tradition in the field of Artificial Intelligence. I do not mean primarily the legendary success on Jeopardy. The tradition dates back to the 50s, when for example researchers at IBM designed Dame Programs that played better than their developers did and developed AI programs to prove geometric theorems.
Stefan Pfeiffer mentions three specific fields of application in his contribution to Blogparade, in which Watson could already assist us in everyday work today. Let us take a closer look at these three:
- "As a personal assistant, it prioritizes the tasks or does them already - such as answer e-mails, arrange meetings or even to strike the right tone of a conversation with others.
- As an expert source, it procures answers and analysis on specific topics. Depending on the sector the system can be fed the relevant expertise. On request it supplies the crucial information to the point and guides the user to the source.
- As Content Manager, it overlooks all of the content that is relevant for the knowledge workers - i.e. emails, presentations, images, etc. It knows not only where they are found, but also learns the context in which they are needed, and provides them accordingly. "
More autonomy or incurred immaturity, Professor?
Here finally Kant comes into play: Would he perceive Watson as useful helper on the way to lead people out of their self-imposed immaturity? At least slight doubts are appropriate, after all he writes:
"It is so convenient to be immature. If I have a book that has understanding for me, a pastor who has a conscience for me, a doctor who assesses the diet for me, etc., then I do not need strive myself. I need not think, if I can just pay; others will readily undertake the irksome work for me." (Answer to the question: What is Enlightenment Chapter 1)
Well, if I reminisce about some presentations about Watson that I experienced lately, I tend that Kant would rotate in his grave. With Watson prevails – to my impression - the industrialized concept of knowledge. As "intelligent materials management system" Watson provides us with the right "part" at the right moment. That may be a step forward but if this leads to "outthink"? I doubt it.
Maybe also because of the crucial aspect that makes the difference between learning and educating is overlooked by many IBM’ers in my perception often overlooked or not developed: Watson might "learn", but he / she / it learns from us how we e.g. prioritize and adapts to us - in the sense of the aforementioned applied artificial intelligence. Admittedly, the published APIs mostly still owe us the actual implementation of this. The major acquisitions of IBM (eg weather channel) show, in my view that the current wave of AI is more in the area of IoT and Big Data. We are only at the beginning of Different. But it will go on rapidly. That ensures the competition between the big players IBM, Google, Amazon, and Microsoft.
What has to happen, so that we people master this development sovereign?
- We need digital science as a compulsory subject in our schools. In fact in the broadest sense of the above-outlined concept of education. Education is just - how Liessmann rightly says - the "program of the Incarnation through the spiritual work on yourself and the world". Said pointedly that means: Latin and digital science are the key subjects for understanding the world. The new teachers are "magisters" i.e. masters, enabling exemplary learning. The learning process itself is here collaboratively.
- Computer ethics become central. In my opinion, an indispensable part of the digitization. Die Roboterethik ist hier nur ein Teilaspekt, aber einer, der fürs autonome Fahren momentan intensiv diskutiert wird. Goldene Zeiten also für heutige Philosophiestudent*innen! (Wenn sie denn auch Informatik studiert haben).
- We do not need to be afraid singularistic visions. Recently there was an interesting interview with Jürgen Schmidhuber on Spiegel Online. There he predicts „in the near futer we [will] have small machines, whose skills correspond to those of a human brain.“ In contrast, we should ask the question of what is human intelligence, or what do people have that machines do just not and cannot have. About that, David Learned wrote an incredibly good book with Tides of Mind.
And certainly much more. Watson, I look forward to the future with you. I hope that there are many partners that use your talents to build a variety of applications, which altogether take us forward. And that the discussion about the sense and nonsense of artificial intelligence is not guided ideologically. Then it could be the beginning of a beautiful friendship. Although currently not in German, only in English. Another challenge for Immanuel Kant.
Here is the book tip: David Learned Tides of Mind. The survey of our consciousness, Published by Ullstein, 2016. Absolutely recommended reading!