"I for one welcome our new computer overlords." wrote Ken Jennings in his Final Jeopardy response today as he conceded defeat to Watson, the IBM supercomputer.
I have been watching Jeopardy with a great deal of interest these past three days as Watson, the IBM supercomputer, has been competing and soundly defeating the two all-time (human) Jeopardy champions, Ken Jennings and Brad Rutter. Watson is being billed by IBM as the first computer that can "think" due to its ability to answer complex questions in natural human language in real time. You can watch a video of how Watson does this here:
However, in truth, Watson does not really think at all. He merely goes through a series of algorithms to come up with the most likely answers and then weighs the probably of each of these answers against each other to determine which one he should choose. A great information gathering machine, yes, but a thinker Watson is not.
However, Watson, has a lot to teach us Jewish teachers. How many of us ask our students to merely be "information gatherers" rather than "thinkers"? DO we merely ask our students "Kitzur Shulchan Aruch" type questions on tests? (As a colleague of mine at Gemara Berura affectionately called it when teachers asked students to merely regurgitate that which was presented in class.) In this case, how are we challenging our students to be more than just a pale imitator of Watson?
Watson can also be trained to identify who spoke to whom in a biblical verse, to list the major causes of the civil war, or present the differing positions in a Talmudic dispute between Abaye and Rava or Rashi and Tosfot. But Watson can't be trained to understand WHY Abaye and Rava argue based on an exact inference from the Mishna, HOW events in the United States led to the Civil War, or WHAT textual basis is there for the differing opinions amongst our medieval commentators on a verse in Tanach. We must teach our students to read, think, and analyze on their own and this is what we must test on as well. How many of us include an "unseen" or application question on our tests where we ask our students to apply the skills or analysis they gained from class to a new text or situation?
I believe that technology can assist in this paradigm shift in Jewish education from loading our students with information to teaching them how to think. (On this topic, Rabbi Gil Student has an excellent blog posting on Technology and Sinai and Rabbi Aaron Ross has a piece on Why Teach Gemara.)
Here are three practical examples.
- Gemara Berura: I have blogged in a recent post about this program. Gemara Berura is a skills based methodology for students to learn how to learn on their own. The strength of this method is that it is not the "Kitzur Shulchan Aruch" approach. The students are trained to figure out what the Gemara is doing instead of answering what the Gemara is saying. In this way, students gain the ability to divide, classify, and connect a Talmudic text using textual cues like the Talmudic sages mentioned, keywords, and nuances of language. Students closely follow the shakla vetarya, the question/answer argumentative style of the Gemara, and eventually learn to "think" like a Talmudic scholar.
- Wikis and Online Discussion Forums: I have also written about this in the past. Students can learn to build their knowledge by sharing with others and through self reflection on these online forums. The asynchronous nature of these activities encourage higher order thinking as students are given the wait time to think before they write and the ability to continue to edit and correct their writing until they are satisfied. Also, reading everyone's response encourages transparency and creativity. There is no one "right" answer so students can't copy off each other and the act of reading other's responses helps kids to develop their own unique ideas.
- Smart Boards and other interactive white boards: Close textual reading is greatly facilitated by the Smart Board or other similar interactive white boards. This does not necessarily require dramatically creative lessons. Rather, merely projecting the text on the board, and then using the highlighters and digital pens to point out important shorashim, words, and phrases can help transform our students into careful readers. For the more creative teacher, the act of putting objects on the Smart Board for students to manipulate can also assist students in thinking through important concepts.
Or new computer overlords have much to teach us. It is my hope that Watson spurs us Jewish educators to train our students to accomplish that which each of us were uniquely endowed with by G-d, to think.

Hello Rav,
ReplyDeleteA brilliant insight into this technological breakthrough. What bothers me is the piece that is missing that none of the modern methods of education (including ArtScroll and Internet Torah) are able to provide and that is the role of the "rebbe". The rebbe is able to educate a talmid not only how to learn and behave but how NOT to learn and how NOT to behave. Books and computers can give a list of instructions but are not able to build a relationship that is the channel used to transfer the massoret from Sinai via Chazal to the learner. Also and just as important, learning with a chavruta under the close observation of the rebbe creates a dynamic that brings life to the whole experience which can nurture true growth.
David
I'm not sure that Watson can't be trained to determine why, how or what. Although I don't think anyone really thinks that computers will ever be able to mimic literally everything the human brain can do, the issue is of pushing boundaries. That's why Jeopardy was such a great test case to train for - there are so many uniquely human things involved in playing the game properly. The same thing with other technologies. Ten years ago OCR technology was awful and primitive. It still is, but great strides have been made since and will continue to be made. I
ReplyDeleteThe purpose of developing AI isn't to undermine humans, it's to aid them. IBM knows it too - and I think you were unfair in describing them as billing Watson as a computer that can think. They are very transparent about what it actually does.
Think of how great a diagnostic tool an artificially intelligent computer can be. Human physicians can't really memorize, internalize and synthesize all medical literature, but a computer may be able to. That wouldn't mean that a Dr. Watson will put physicians out of business - they will help human physicians, just as technology already helps them.
IBM admits that Watson can't think. See the following:
ReplyDeletehttp://www.ibm.com/ibm100/us/en/icons/watson/
IBM’s scientists have been quick to say that Watson does not actually think. “The goal is not to model the human brain,” said David Ferrucci, who spent 15 years working at IBM Research on natural language problems and finding answers amid unstructured information. “The goal is to build a computer that can be more effective in understanding and interacting in natural language, but not necessarily the same way humans do it.”
The challenge for teachers in an age of Watson is to make sure we design lessons that let our students think because if we just ask them to memorize and "spit back" then how are they any better than Watson? Actually they're worse since Watson is really really good at that as evidenced by his Jeopardy performance.