Share and discuss this blog



Announcing Schank Academy Logo

Welcome to the future of learning - the way it was always meant to be.

Courses start June 2017. Enroll today.


Monday, November 20, 2017

Sexual Harassment Training? You must be kidding me

I feel inspired to tell an old story because of a story I read in the Washington Post yesterday:

What’s the point of sexual harassment training? Often, to protect employers.


Here is one paragraph from that article:

Unfortunately, there is little evidence that training reduces sexual harassment. Rather, training programs, along with anti-harassment policies and reporting procedures, do more to shield employers from liability than to protect employees from harassment. And the clearest message they send is to the courts: Nothing to see here, folks.

About 20 years ago, my company had a contract for building training (it wasn’t then called e-learning) on the computer for a company which was at the time one of the biggest technology firms in the world. We built a lot of things for them that they liked a great deal. Then, they asked us to build a sexual harassment training course. My model for building courses was, and is, to create as realistic a simulation as we can so trainees can try and fail. Then, and now, I believed that we learn by doing and cannot learn from being told. But, this contract presented a problem.

What could we build? The obvious choice was to build a fictional attractive woman and then put the student in a situation where he might want to harass her in some way. This idea was so profoundly stupid that there was no way we would build that. We needed to think of something else.

We often hear about training that works great to deal with compliance issues. Here is something  I found in the National Review on diversity training:

In the United States, it’s just a fact that you’re supposed to arrive on time to your appointments. According to materials from a diversity-training course at Clemson University, it’s culturally insensitive to expect people to show up on time because “time may be considered fluid” in some cultures. Clemson’s “Diversity Benefits for Higher Education” initiative — which cost the school more than $25,000, according to Campus Reform — presents its participants with slides featuring hypothetical scenarios, and asks them to select the correct, culturally sensitive action from a list of options. “Alejandro scheduled a 9:00 a.m. meeting with two groups of visiting professors and students from other countries,” one of the scenarios states. “When he arrived, he found that the first group had been waiting for fifteen minutes.” “The second group arrived at 9:10 and wanted to socialize first,” it continues. “What should Alejandro do?”  

The slide then lists three options: 

1.  “Politely ask the second group to apologize.” 

2.  “Explain, ‘In our country, 9:00 a.m. means 9:00 a.m.’” 

 3.  “As the meeting organizer, he should recognize cultural differences that may impact the meeting and adjust accordingly.” 

The correct answer, according to the slide, is option three.

The correct answer, according to the slide, is option three. “Time may be considered precise or fluid depending on the culture,” the slide explains. “For Alejandro to bring three cultures together he must start from a place of respect, understanding that his cultural perspective regarding time is is neither more nor less valid than any other.” Sorry, but — nope.

Sorry to say that that is the kind of “training”  that our client expected. But, sorry, I won’t do it that way. You can’t learn by being told the right answer.
But, you can learn from doing so we had to come up with some doing. 

We decided to “train” students to do something that they would never have to actually do in real life. We trained then to be sexual harassment adjudicators. Cases were presented to them, (both sides of the story) and they had to determine, by consulting the company’s guidelines, whether this was a case of sexual harassment, and if so, what to do about it. It was an interesting course and the students really enjoyed it. Moreover, the company really liked it. They had expected nothing like it. It was fun and it seemed to  work.

Except, well, our point of contact at the client was a very attractive woman. She dressed in a way that called attention to herself. She was also a friend of mine (which is how we got the work.) At the end, she told me that her two male bosses (whom I knew because we were working with them) had been harassing her the entire time that we were building this course.
Compliance training is “often to protect employers”? No. It is always built to protect employers.

This from the Los Angeles Times:



Former Google employee James Damore was supposed to come away enlightened by his diversity training, armed with a newfound sense of empathy for colleagues who did not look like him, a white male.

Instead, the software engineer was so enraged by the experience he decided to write a now-infamous 3,000-word memo on a flight to China railing against Google’s “ideological echo chamber” and arguing that women land fewer tech jobs because of biological differences.

“I went to a diversity program at Google and … I heard things that I definitely disagreed with,” Damore, 28, told Stefan Molyneux, a libertarian podcaster and author. Damore said he had some conversations at the program, but “there was a lot of, just, shaming — ‘No, you can’t say that, that’s sexist’; ‘You can’t do this.’… There’s just so much hypocrisy in a lot of the things that they’re saying.”

Much of what companies build for training purposes is just meant to tell employees the “truth.” But, it turns out you can tell people anything you want to tell them, but they are likely to forget it, they may often disagree with it, and you aren’t going to change their prejudices or natural inclinations by talking at them. But, we are really used to this kind of “education"  (which I like to refer to as TT&TT tell them and test them) so that is what people build.

The issue here is not about compliance training, which is almost always nonsense, but about education itself. We need to get past TT &TT. We need to do it in corporate training, and we need to do it in school. Can we?  Only if we realize that “one size fits all” is not an educational philosophy. People are different and we need to deal with those differences. We don’t have to accept bad behavior, but neither will it go away by TT&TT. It won’t even go away if you build interesting simulated experiences.

What can we do? One method we use is Observe and Critique which we have used for years. A good example can be seen here: (It is training  we built to teach doctors how to talk to their patients about cancer.).



The difference here is obvious. Doctors would like to learn how to talk properly to patients, one would assume. What you cannot assume, is that trainees want to learn what you are teaching them. If they don’t want to learn it, you can be sure they won't learn it. 

Wednesday, September 27, 2017

If you build it they won't come; making the schools change is hard

Yesterday there was announcement about the urgency of teaching computer science in K-12:


I don’t know about the “urgency” but I can say that we have already built what the schools need: a year long all day, everyday, intensive online, mentored, deep dive into learning to program:


All the companies that say they want to help need to know where the problem lies. If you build it, they won’t come. 

How can you get the schools to launch something like what we built? You can’t. You would have to eliminate one of the sacred liberal arts subjects that we all endured in high school. Throw out history or literature or chemistry? Not happening. Until Harvard stops requiring these subjects of its applicants, no one can make a single change.

Apart from the sacred subjects we also have the sacred notion of a “course.” A course is one of five students take at the same time, and therefore needs to be taught one hour each day. You can’t really learn to program except by programming a lot. Listening to the teacher talk (who probably does not know how to program anyway) and then coding at night for an hour will not turn a kid into a programmer.

As it happens, I met Monday with a college (who I will not name) that understands the urgency of cybersecurity and wants their students to be able to learn about it. So, they have a course or two to offer. We have another deep dive (a six months all day intensive online mentored learn by doing course.) 


Can this school offer it? Of course not. The faculty would never agree to such a thing. Professors want to teach 3 hours a week. They won’t agree to teach more hours and none will agree to the idea of a student doing only one thing in a semester.

So, while I am heartened by the idea that the government and industry would like to help the schools change, it is important to remember that the schools do not want to change.

All that money would best be spent on building new schools that don’t have an embedded faculty with vested interests. And, until we abandon Common Core (ironically Bill Gates’ contribution to education), we will not have computer science in the schools in any serious way. 


We need to get rid of the subjects we teach in school and the way school is structured. Only then can we introduce real change. And while we are at it, we must make Harvard come up with requirements that do not reflect Roman ideas about liberal arts education.t come

Tuesday, July 18, 2017

AI requires an understanding of intent: a look at an old movie




I was watching a movie from 1934 called “Chained.”  Clark Gable plays an American rancher whose ranch is in Argentina. They are rivals for the same woman. Gable’s  rival is a richer and classier man who is a shipowner. They meet in New York (together with the woman they are fighting over) and have the following conversation:


do you ship any of your animals up here?

yes, some cattle

I hope you use our boats

no your rates are too high

well, we’ll have to see about that

my animals are healthy, they can travel on a slower boat; less money

good economics; where did you learn about that?

I was at Yale

I’m Harvard


This conversation tells one all there is to know about AI, education, and learning (my three favorite topics.)

First AI. How would you get a computer to respond to

good economics; where did you learn about that?

with I was at Yale?

First, this is not an answer to the question. Also, the question isn’t what it seems to be. Where did you learn not to waste money? is a fairly obnoxious question. The word “economics” makes the question sound as if it were more than it was. But, one doesn’t ask someone who has said something simple (I try to save money when I can is all he was saying after all) about where he has learned it. Unless of course you are talking to a four year old. This is a remark made by someone who thinks he is much better than the person to whom he is speaking. 

I was at Yale is not an answer to where Gable learned this at all. It is an answer to the underlying snooty question that is really being asked, which is more or less I guess you aren’t as dumb as I thought you were. The answer I was at Yale means I am fancier than you thought I was.

I’m Harvard is a response that says that we are in the same class after all. I feel better about talking to you now.

How would a computer understand the power games going on here? How should it understand that Harvard and Yale are actually answers about social class and that this conversation is in no way about economics nor is it about any economics courses that Gable might have taken at Yale. It is about two people sizing each other up, which is something that often goes on in a conversation where two powerful people meet, and especially one where a woman they are fighting over is at the table.

So here is a translation of this conversation that most people who were watching this movie would have  subconsciously understood:

how do you do your business?

can I make money on you?

no your rates are too high, I’ll explain 

my animals are healthy, they can travel on a slower boat; less money

well; you aren’t stupid, but do you have any actual education? Are you in my social class?



I am as good as you are

So you are.


Can a computer do any of this?  Of course not. 

In order to understand what people’s real intentions are and what they really mean by what they say you have to understand a great deal about the world in which we live. In fact, since is a 1934 movie after all, many younger people might have trouble understanding it. If you don't see Harvard and Yale as class markers, you would miss it. In 1934, Harvard and Yale were the ultimate class indicators. Today, this is still true but in a different sort of way. 

Today, we would see  I’m Harvard as being an odd way to talk, not as a way of being snooty and saying I am as good as you are. But we also, in today's world, understand that anyone who goes to Harvard must be very good at test taking, and must have been an all A student in high school. This was not true in 1934. 

Elon Musk can fear smart computers all he wants, but until computers can glean the underlying intent of a sentence on many levels he hasn’t much to worry about. Modern AI is about key words and search not about ideas. The main ideas in this dialogue are never actually mentioned, so searching key words won’t help much. The shipowner is saying, more or less: let’s find out how good you are because I am sure I am better than you and therefore better for this woman we are fighting over. The rancher’s response is I am as good as you are. In fact the movie has the woman going back to Clark Gable because of this very scene. She had previously said she wanted money and social status and Gable proved he was his rival’s equal.

Where can you find all this? Not in the actual sentences uttered, that is for sure. Understanding is about context and inferences and intent. One must figure out other people’s goals and plans in any conversation. Which of today’s ‘AI’s’  can be in a conversation and figure out just the right thing to say to undercut the other person?

How do we learn to do all this? Might we learn it at Yale? You would, in fact, but probably not in class. Real learning, the kind that forces us to re-think what we are doing and look for the underlying intentions of those with whom we are engaged, takes place in the dorms and after class. The teacher can babble at you all you want but you still have to talk to people, sound intelligent, figure out where they are coming from, and attempt to engage them and perhaps convince them. This is learned at Yale but it is not taught at Yale.

In fact, very little of that we need to teach people and what people need to learn in order to understand the world is learned at school, its learned from experience and reflection on that experience.

We have AI wrong, learning wrong, and education wrong. Other than that we are doing fine.


Tuesday, July 11, 2017

Take Deep Dive when you want to learn something


The latest fad in e-learning is usually something silly, but the current fad of nano-learning (look here if you want to know more about this ridiculous fad:  ( https://elearningindustry.com/6-ideas-help-create-nano-learning-15-minutes  ) drives me especially crazy. Why? Because it is the exact opposite of the solution to the fundamental problem we have in education and learning. 

When we learn something and feel that we really know that thing, something is always the case. We have taken a deep dive into that thing.

What do I mean when I say “deep dive?” You have taken a  deep dive when you immerse yourself in something, obsess about it, look at everything through the perspective of that thing and make yourself an expert. 

Our current system of education makes deep dives very difficult to do. At this moment I am thinking about this because we have built a 6 month long deep dive into cybersecurity for the Pentagon. Trust me, it takes at least that long to really understand cyber warfare. 

So, now we have to roll this course out. When I was first contacted about building this course, the Pentagon’s plan was to roll it out to soldiers. I wondered how many soldiers had the “hacker mentality” which to put it briefly is about breaking into things not about following rules.

But I am now starting to understand that while this may not be a great plan but it might be the only plan. Why?

Businesses say that they are worried about cyber defense and they certainly should be. But, businesses will not take someone off the job and send him or her to school for 6 months. They may be worried about cyber but they aren’t worried enough to do that. We will have to build 2 day long courses, or maybe even 2 week long courses in order to get corporations to send students.

When I realized that this was probably the case, I also realized that this has always been my problem with the people in charge of training at corporations. Of course they all want nano-learning now. They never want to send people to training (and now complain about them having to spend even an hour on it). They send people to training because they have to, but they have always wanted things to be fast. 

When I first started working in corporate training, I was fascinated by that fact everything that needed to be taught was taught in a one week long course. Why one week? Because they flew people in to the training center on Sunday and flew them home on Friday night. Online learning has only made this problem worse. Now that they don’t to fly anyone anywhere they can make the courses even shorter.

Curiously, the one week course is identical in length to college courses which usually meet three hours a week for 14 weeks, about 40 hours. How is it that everything you learn in college takes 40 hours also? This easy enough to explain. No deep dives.

If you want to do a deep dive into something in college, you need to wait for a PhD program or you can sometimes finagle a senior year where you can devote one semester to doing only one thing. But, this is very hard to do. The education system actually prevents deep dives rather than encouraging them.

School is nearly always shallow. It is obsessed with covering the material rather than with mastery of something. So, we memorize short bits of information, the Pythagorean theorem, how to balance a chemical equation, who won the battle of Hastings, who the main protagonists are in A Tale of Two Cities. We do this because we are told to do it. And, not surprisingly, we forget most of what we have memorized after we take a multiple choice test.

Lately, this kind of education is justified by the idea that one is teaching “critical thinking”  (although it used to be justified by covering things every educated person must know.) “Critical thinking” is now used to justify doing things the way they have always been done. 

As an example, The United States Military Academy (USMA) offers thirteen possible majors. Now, just let your mind wander for a bit and try to imagine what those majors might be. USMA is meant to train future officers for the Army. Students know what they have signed up for and the Army knows what kinds of graduates it wants to produce. My guess would have included majors in warfare, weapons, strategy, leadership, discipline and such. Not that I would have thought that there would be majors at all. USMA isn’t a liberal arts college after all. They needn’t cover academic subjects nor do they have to teach in the style of Harvard and Yale. Unlike Harvard and Yale, entering students know what they have signed up for and what they will be doing when they graduate.

I guess I was just dreaming. Our current sense of what goes on in school is dominated by academic subjects. You can major in history at USMA. Really? A course or two sure. But Army officers who were history majors? Couldn't we teach them stuff they would actually use?

My son told me he was going to major in history after being in college (at Columbia) for a week or so. I told him to get on the next plane and come home. Why would I do that? Because college ought to be a place where one takes a deep dive into something, coming out prepared to do something in the real world that interests one greatly. In my son’s case, I knew what he wanted to dive into since he was eight years old. He was obsessed with subways. But, no surprise, there was no subway major available. I told him to figure out how he could create one. (It all turned out fine. He is still playing with subways (in Los Angeles. He is 42 years old and Chief Innovation Officer for LA Metro.) Why wasn’t someone talking to him about his interests and helping him make a deep dive into subways? Because colleges don’t work that way of course. But shouldn't USMA work that way?

But, you can be a history major at USMA.  I couldn't believe my eyes when I read this:


The Department of History's Mission is to educate, train and inspire the Corps of Cadets through the discipline of history so that each graduate is a commissioned leader of character committed to the values of Duty, Honor, Country and prepared for a career of professional excellence and service to the Nation as an officer in the United States Army.

As Army officers, West Point graduates will perform a broad spectrum of missions vitally important to our nation's security and interests. They must be intellectually and professionally prepared to face these challenges in an uncertain and dangerous world inhabited by peoples of different languages, religions, and cultures. The Department of History contributes to cadets' intellectual and professional development by imparting historical knowledge, an appreciation of history, and critical thinking and communication skills.

Officers who are critical thinkers challenge accepted wisdom in the search for truth and justice. They are open-minded and able to make independent and informed decisions. They reject simplistic answers that suggest the existence of a black-and-white world; rather, they accept the ambiguity associated with most human endeavors and seek the best solution rather than a single "correct" one. The study of history encourages critical thinking by requiring cadets to:
formulate critical questions;
conduct research by gathering and prioritizing information;
analyze information within the broad context in which it appears;
interpret and synthesize information;
derive reasoned, evidence-based conclusions;
assess and adjust their conclusions as conditions change or new information becomes available.


Wow! Of course, this major is justified by the notion of “critical thinking.” Although it is fairly obvious that one learns to be a critical thinker in any area of life that one dives into, the study of the liberal arts are often justified by that expression. 

You want to be an English major at USMA? No problem:


We prepare cadets to be outstanding communicators and adaptable critical thinkers who can synthesize concepts; appreciate diverse cultures, ideas, and forms of expression; and assess clearly the implications of complex ethical questions.


Because the Army needs officers who can talk about the major themes in Shakespeare. Oh, and they need critical thinkers. 

I was surprised to find that you can be a math major at USMA. I was a math major. It prepared me for nothing. From the USMA math site:


Our purpose is to provide each cadet the opportunity to gain the mathematical education essential to progressive and continuing development throughout a career as a Regular Army officer. Emphasis is placed on achieving intellectual discipline, mastery of reasoning, understanding of mathematical concepts, skill in practical applications of mathematics and appreciation for the role of mathematics in the military.


What is the problem here? Don’t we need officers who can reason? Of course we do. Will they be using the Quadratic formula when they reason on the battlefield about how to confront the enemy? One wonders what “mathematics in the military” might mean? I guess one might need some math to figure out how to fire a long range missile or to do cryptography. Is that the kind of math they teach to their majors? If it is, wouldn't all officers need that?  

The issue here is allowing students to take a deep dive. That is not the same as a major. The very concept of a major has more to do with the needs of the faculty than the needs of the students. Students are required to become “well-rounded” in a field. But the real requirement is that every professor wants people to take the course they teach so computer science majors (for example) must learn theory they will never need because computer science faculties always have theoreticians on there faculty. (The reason for this is that theoreticians cannot find employment outside of the university.)

And sure enough, if we go back to USMA we find when we look up the computer science major:

The Computer Science (CS) major at the US Military Academy develops fundamental competency in theoretical and technical areas of computing, as well as a characteristic style of thinking and problem-solving.


Moreover the key issue isn’t what USMA wants or needs and can be found on the same page:

The Computer Science program is accredited by the Computing Accreditation Commission of ABET

In other words, accreditors have decided what army officers must know about computer science. And you can be sure that  ABET has theoreticians as part of the accreditation team.

So, the very opposite of a deep dive, is typical of most majors in anything because a committee that wants to satisfy all the faculty has determined that a major should skim the surface of the many sub-parts of a discipline.

In real life however, deep dives are the norm. For example, piano players must take a deep dive into piano playing in order to be any good. They have to practice endlessly and continually try more difficult things. They needn’t memorize the six principles of piano playing or major in piano. They just need to dive in and keep going.

This is true of any profession. Doctors need to dive into medicine. So of course, one would think that medical schools allow that. But they really don’t. Doctors have to memorize the names of bones and take courses in anatomy. They do not really dive into medicine until they are interns and then their lives, if they are really committed, are a continual deep dive into new procedures and findings. The reasons for this in medicine are the same as those in computer science. Shouldn’t every doctor know a little about everything in medicine? That sounds right to faculty who want to teach their specialties, but that attitude prevents deep dives (which are relegated to internships in the case of doctors.)

We are concerned here with what it means to restructure educational practice into deep dives and get away form the smorgasbord approach to education. Colleges were structured the way they are today a very long time ago. We will discuss this later. For now we want to consider an online course we have built that makes clear what a deep dive is and should serve as a model for education at all levels. 

Cyber 

Our cyber course is made up of 17 tasks plus a very difficult capstone. Each one follows the story line of the one before and builds on what was learned there. Before we begin to discuss that course in any detail we want you to hear from the first student who took that course. She is someone who never graduated high school and was working as a massage therapist when she decided to test out our course:


Task 3 made me a little angry.  The task was designed perfectly.  It had a lot in common with the second task but it required you to kind of approach it more analytically to figure out exactly what was going on and I figured it out after I conducted four little experiments that necessitated a fairly analytic approach to the problem, and is just really fun to do.  But it made me realize what a huge disservice our educational system is doing with the way analytic thinking is usually measured, because it's usually assumed that doing math measures your capacity to think analytically, right?  So I could never take any interesting science or technology classes earlier in life because they always had some stupid math prerequisite I refuse to fulfill, because I absolutely hate math.  Like I don't even know my times tables, that’s how much I hate it.  So I resigned myself to a life outside of the sciences, but I really love the process of working out a solution to a puzzle in a thoughtful rigorous systematic way.  So I ended up getting really angry once I realized how much of the course of my life had been determined by stupid math prerequisites that could have been fulfilled in other ways, like, for instance how I went about solving this task.  And it also made me really thankful that a course like this one exists where it matters more what you can do than whatever your resume happens to look like, and that's pretty amazing for people.



She never graduated high school but she is becoming a very good hacker which requires serious attention to detail and reasoning to figure out what is going on in a world that is intentionally hidden from view.  Here she is talking about why she avoided computer stuff until now:


My mom is a Professor of Computer Science and I’ve always had a proclivity new for playing with computers.  I would pick up computer skills quite quickly.  I was spending days writing exquisite regular expressions as a teenager to help automate and clean up and format text, scanned in OCR just so I can annotate what I was reading more easily.  I was a 9 year old. I catalogued all the glitches I could find in my favorite video game and made a website to share all the fun bugs I found.  But it never even occurred to me to pursue anything computer related as a career because I found the people in the computer science department excruciatingly boring.  Subsequent experiences with IT staff and people that enjoyed programming did nothing to mitigate that impression.  What I wished somebody had told me as a teenager was that the most interesting people who play with computers are hiding in computer security.  I wish somebody had told me that you can get into this field without being a black hat first.  There’s good hackers not just black hats.  And you still can be a good hacker even if you’re too nice to troll anybody.  I probably would have gone into this field fifteen years ago if I realized that’s where all the interesting people who like tinkering with computers are.  And if a daughter of a computer scientist doesn’t know that, what chance do other women with the same proclivities have?  I never would have realized any of this. But I just happened to know somebody who said “I think you might be good at this.  Don’t worry these people are really different from the normal CS people.”  So I think a big part of the problem is the PR problem for computer science.  


Now, she is not a “normal” kid. In fact she would be classified as being on the autism spectrum. But the interesting thing about that “spectrum” is that it is used to take kids who are not compliant and get them out of the classroom. Teachers don’t want students who refuse to do what they have been asked to do.

This “spectrum” didn't exist when my kids were in school. They both hated school and were difficult to deal with. My son refused to do his reading homework when he was in the third grade. The teacher sent for me. I asked him why he wasn’t doing it and he said it was boring and tedious. I looked at it. He was right. He had to read random paragraphs and then answer  multiple choice questions about that paragraph. 

When I met with the teacher she agreed that the homework was boring and asked what I suggested. I said my son like to read. He could be assigned to read a book. (Or even more radical, he could choose a book he wanted to read and report on it.) 

The teacher thought this was a fine idea and my son was happy. But, a few weeks later I was called in to see the teacher again. She said they had to abandon my “read a book” idea. Why? Because my son had been in the advanced reading group. The other kids in the group were complaining that reading a book was too hard. They were in that group because they were good at reading short paragraphs and answering questions. They didn’t want to do the “deep dive” that reading a book entailed.

As long as we see learning as a quest for grades,  credits, and “coverage”,  most students will prefer doing as little as possible to achieve that and they will not fight back against the system. To put this another way, our education system is rigged to be simplistic, and most students prefer it that way. Except the students who have a real desire to do something want to take the deep dive. They also want to be able to choose which deep dive to take. (My son would have chosen to read every subway book he could find; instead he had to read Dick and Jane.)

Deep dive learning means exactly that — allowing students to take a deep dive into what they want to do and then letting them go at that for as long as they want. What bad would happen if we allowed that? Well they might not learn history or literature. But as we see from USMA the entire point of learning these subjects is to teach critical thinking which could certainly be taught from a deep dive into any complex area.

Take a deep dive and learn how to do something complex. Shouldn’t that be our approach to education instead of the one size fits all approach we have now?

You would think that corporate training people would understand this, but they do not. They want courses that are an hour long. Did I say an hour? I meant five minutes.


Take a deep dive and learn to do something real. Schools can’t allow this because their structure is immutable. That will have to change and change soon if we want to produce students who are job ready.