that's random

Has the term AI become meaningless? How about MI instead

Ian Bogost writing for The Atlantic says that in too many cases today “artificial intelligence” is just another name for a fancy computer program. I don’t see it that way. I know from experience that what most data scientists are building is entirely different from what rank-and-file software developers are building. We use different tools and different approaches. And the data-driven learning algorithms we deploy at their best solve an entirely different class of problems than regular computer programs do.

Personally I like to call what we build “machine intelligence” rather than “artificial intelligence” because machine intelligence is really an alternative kind of intelligence, not an artificial version of human intelligence.

No, it’s not “Making computers act like they do in the movies” as Bogost quotes AI researcher Charles Isbell. That is too glib indeed. Why not let machines do what they do best rather than just serve as poor imitators of humans?

Part of what makes “artificial intelligence” feel a bit underwhelming is that we’ve barely begun to see what we might achieve with machine intelligence. Yes, self-driving cars are pretty amazing. I don’t have one myself (can’t afford a Tesla, darn) but I do adore the parking sensors on my SUV. They allow me to navigate around the dangerously-placed porch jutting out by the attached garage set back to the rear of my house. If I didn’t have them I probably would have hit the porch at least once already. The car can parallel park itself too but I’ve only tried that once, before I bought the car, with the salesperson sitting next to me.

I have faith that we are going to see many more amazing machine intelligence capabilities come out, as startups and big companies start focusing on vertical artificial intelligence in specific domains rather than continuing to build out horizontal machine learning capabilities for use by data scientists. Vertical AI (or MI) is tough. That’s where you have to get domain experts and data scientists together and figure out how to encode domain expertise and capabilities into machine learning models. It’s tough and slow work. I know. I’ve been doing it for a few years now in the temporary workforce management space. We’re beginning to see the payoff though, and that is truly exciting.

If you want to hear about it, I’m going to be at VMSA Live in Phoenix in early April talking about Machine Intelligence in Talent during the Executive Gateway session on Wednesday, April 5th. If you’ll be there, stop by.

that's random

Should everyone learn to program? And by everyone I mean women

This is something I’ve been wanting to address, because I care about encouraging women to enter and succeed in STEM fields, because I am/have been a programmer, and because there’s a lot of angst around this. No, the discussion is not specifically about whether women should learn to program, but it is especially pointed and sometimes painful for women. Girls are not usually encouraged towards STEM careers and even less towards straight computer science. When women do make it into STEM fields, they may find themselves marginalized or otherwise stressed by the overwhelming dominance of men in technology.

Background reading:

On one hand, I think the answer is easy: if you need to know how to program to progress in your career and achieve your goals, then just do it. I went back to school after completing a bachelor’s degree in economics and philosophy so that I could gain programming skills. I did an M.S. in statistics but spent all my electives in the computer science department, because I knew that if (in the early nineties in Silicon Valley) I could demonstrate programming skills, I would be welcomed into the job market. And I was. Then I found many years later, that statistics + programming + domain knowledge = data science = $$$. Score!

On the other hand, I see serious obstacles for girls and women here. There are a set of reinforcing messages that girls/women receive that keep them from just doing it:

  1. Math (… computer programming … physics … etc) is damn hard.
  2. If you don’t have the brain for math (… computer programming … etc), don’t bother. It’s too damn hard!
  3. Girls don’t have the brain for math (… computer programming … etc).

Our culture thinks that natural aptitude matters more than hard work in figuring out subjects like math and programming. This is not the case in other countries; in Asian countries it is thought that academic achievement including in math and science is primarily due to hard work. It is the performance mindset in action rather than the growth mindset, which says that hard work is almost everything. There’s just a short step from the performance mindset that dominates in the U.S. to thinking that certain groups have natural advantages in learning such subjects and that certain groups have natural disadvantages.

An inevitable outgrowth of this is generalizing from individual cases to a group:

How it Works (xkcd)

I face this pretty regularly even though I work in education, which has a decent balance of women and men, if not lopsided in favor of women in non-tech areas. I recently went to a “big data” internal meeting where I was the only female in attendance. This is not at all unusual for me in my position–when we’re talking tech it’s usually mostly or only men, besides me. I am aware in such meetings that when I say something I am representing female techiehood. If I say something dumb, I have shown that all females are technologically clueless. But if I say something smart I imagine people (/men) may be thinking “well sure, this one woman knows a little bit but she’s not a representative case.”

That sort of bothers me, because I know in some ways I come across as not a representative case for women’s capabilities at large.

I am, however, a representative case of why more women don’t study computer science in college. Despite having taught myself BASIC on my dad’s Apple IIe in middle school, I refused to take Intro Computer Science as one of my distribution requirements in college after hearing from (female) friends that the class was killer. I didn’t believe I could hack it. I came back and took it in my master’s program. To my shock, it was not only easy, it was fun. I got an A+.

If I consider where we need to aim, I’m thinking it’s not at the idea of whether girls are good at math/programming/etc or not but rather from this idea of whether such subjects require natural facility or not. They do not. They require a lot of hard work, for everyone, no matter their gender or their race or their obvious natural facility for things technical.

I do not want to dismiss the very real social challenges women meet in male-dominated environments. I cannot speak much to that as I’ve never felt unwelcome or harassed among male-dominated teams. Quite the opposite: I find working with men exciting.

So ladies/women/girls, dive in. The water’s cold and the current is strong: you’ll get a good workout. Enjoy!

that's random

Desultory musing* about mashed-up selves

Trying to find yourself is a staple of the self-help literature, along with the striving for authenticity and building up your self-esteem. I probably wrote about authenticity and how you needed to practice it in my book** because way back in 2007, I thought it was a good thing, a necessary thing.

Now I’m convinced that’s all wrong. The self that matters isn’t some tightly defined, self-loving, individuated thing in the world. The self that matters is the mashed-up self, the networked self — the self made up of relationships and experiences and interactions and ideas. It’s way bigger and more powerful than the un-networked you.

These are some ideas I want to explore: combinatorial creativity, connectivist learning, the third person perspective in the creative process, and self-transcendence. What all these have in common is they all overturn the idea that the individuated self is primary.

Writer and artist Austin Kleon on how we are mashups:

We can pick our teachers and we can pick our friends and we can pick the books we read and the music we listen to and the movies we see, etcetera. You are a mashup of what you let into your life. [via Maria Popova]

Maria Popova on networked knowledge and combinatorial creativity:

Which is interesting, recognizing not only the absolute value of content but also its relational value, the value not just of information itself but also of information architecture, not just of content but also of content curation….

The idea that in order for us to truly create and contribute to the world, we have to be able to connect countless dots, to cross-pollinate ideas from a wealth of disciplines, to combine and recombine these pieces and build new castles.

This relates to how we conceive of ourselves. Are we distinct individuals with hard boundaries? Or are we somehow only ourselves when you consider how we fit into a network of experiences, people, and knowledge?

Robert Fritz, author of The Path of Least Resistance, is someone who might be called a creativity guru but instead of that I would call him a creation guru. He doesn’t write so much about being creative as about actually creating. In his view, the self that we want so much to develop and pin down should be set aside:

Don’t try to define yourself, instead, suspend the question. That gives you a better lens by which to create anything you want to be.

Here’s what Fritz discovered when he tried to help people become more creative by improving their self-image:

Questions of identity got in the way of the creative process, even when people thought well about themselves.  It took years to come to understand this.  Everything pointed to a fairly simple and, yet, revolutionary pair of insights.  The first was that what people thought about themselves, good, bad, or indifferent, wasn’t going to change.  All notions of self-esteem training are predicated on the idea that people can change how they see themselves.  This is one reason they don’t work.

The other insight was just as major.  It is that your view of yourself has no place in the creative process.  Simply put, the moment you make your success or failure about you, that’s the moment you can’t learn what you need to learn, experience what comes with the creative territory, and keep your focus where it needs to be, on the outcome you are working to create, and where you currently are in the creative process.

In his books, Fritz suggests that creators need to use a third-person perspective that takes themselves out of the equation rather than the first-person which makes creation all about the self, the I, the creator herself. Creation doesn’t grow out of some authentic, independent self. It launches from a networked self which is almost like no self at all.

Maybe the reason that thinking too much about our identities as distinct individuals stops us from creating is because creation comes through mashing up, through navigating networks of people and knowledge and ideas, not from the perspective of one isolated node in the network. The node alone is useless.

So we need to stop thinking so much about our individual selves — we need to transcend ourselves. Interesting that  some of the most satisfied people combine a love of the new with persistence and self-transcendence. These seem like exactly the traits you’d need to succeed in a networked world. Neophilia (novelty-seeking, love of the new) draws you to new ideas, new people, and new experiences, giving you more material for the mashup that is you and the mashups you create. Persistence keeps you from being merely a dilettante, flitting from one new thing to another. And self-transcendence stops you from thinking that it’s all about you.

It’s really freeing to realize your self alone is this puny, incompetent thing whose self-love or self-loathing matters not a bit. It’s your networked mashup self that matters, that’s capable of doing and creating great things.

* With credit to JP Rangaswami for his “musing lazily about” series. I like to be able to wander around a topic without reaching any conclusions or forcing it into some structure that might obscure the evolving ideas.

** Which shall remain unnamed and unlinked because I’m so beyond what I wrote then even though I feel like there were some really great ideas that I’d like to expand upon and refine. E.g., busy vs. bursty.

that's random

Big ideas require social connection

Neal Gabler in The Elusive Big Idea thinks online connectedness implies a “post-idea world”:

It is certainly no accident that the post-idea world has sprung up alongside the social networking world. Even though there are sites and blogs dedicated to ideas, Twitter, Facebook, Myspace, Flickr, etc., the most popular sites on the Web, are basically information exchanges, designed to feed the insatiable information hunger, though this is hardly the kind of information that generates ideas. It is largely useless except insofar as it makes the possessor of the information feel, well, informed. Of course, one could argue that these sites are no different than conversation was for previous generations, and that conversation seldom generated big ideas either, and one would be right.

But the analogy isn’t perfect. For one thing, social networking sites are the primary form of communication among young people, and they are supplanting print, which is where ideas have typically gestated. For another, social networking sites engender habits of mind that are inimical to the kind of deliberate discourse that gives rise to ideas. Instead of theories, hypotheses and grand arguments, we get instant 140-character tweets about eating a sandwich or watching a TV show. While social networking may enlarge one’s circle and even introduce one to strangers, this is not the same thing as enlarging one’s intellectual universe. Indeed, the gab of social networking tends to shrink one’s universe to oneself and one’s friends, while thoughts organized in words, whether online or on the page, enlarge one’s focus.

[via Stowe Boyd]

But big ideas have always rested on some sort of social connection across people and they haven’t always required the written word. Big ideas don’t issue forth from the head of one really smart person, working alone, reading, reading, reading then Eureka! No, they arise from the synthesis of multiple ideas and practical knowledge that together are greater than the sum of their parts. What’s required for this? That people be in some sort of social connection with each other.

Big idea: Cliff dwelling

Last weekend, we visited Mesa Verde, an archeological site and national park that shows how the Ancestral Puebloans lived in the time period from roughly 550 to 1300 A.D. While the cliff dwellings themselves inspired awe (and sometimes required special tickets and patience to tour), it was more interesting to me to see the progression of architecture at sites along the Mesa Top Loop Road. Here’s where you could see the Puebloans build towards the big idea of cliff dwellings. From 550 to 750 A.D., they mostly lived in pithouses on top of mesas but sometimes in cliff alcoves, then progressed to adobe houses clustered into villages. By 1000 A.D. they had developed stone masonry techniques for constructing buildings two or three stories high with 50 or more rooms. It was about 1200 A.D. that they moved their buildings into cliff alcoves, a spectacular form of architecture with practical and aesthetic benefits.

Who came up with the big idea to build a multi-story stone dwelling in the shaded and protected alcove of a rocky cliff? It certainly wasn’t one person acting in isolation and doing a lot of reading and writing. This big idea did not rest upon the written word and lots of deep thought but rather on social knowledge and practice of stone masonry construction and on ideas and experience about where the best place to live was (mesa top or cliff alcove?). If the Ancestral Puebloans did have Twitter, I could imagine them using it share tips about how best to construct a beautiful and sturdy multi-story dwelling, sited for protection and convenience. Most big ideas rest upon a wealth of little bits of knowledge that don’t need to be written up as a journal article or book.

Many ways to connect

How we connect differs depending on where the people we want to connect with are. In the 17th and 18th centuries, French intellectuals gathered in salons. In the 20th century in the U.S. and other western countries, connecting through higher education channels and journal articles and at government-funded research centers was a good bet. Now, some people connect online using tools like Twitter and Facebook and Google+. This doesn’t necessarily replace long-form writing but it does complement it — making the generation of big ideas more, not less, likely.

that's random

The significance of t-tests

Ph.D. Topics : Statistics

No, I’m not talking about statistical significance here; I’m talking about practical significance.

The first statistical significance test an intro stats student learns is usually a t-test to test differences between group means. If she goes on to use statistics, she may never use a t-test for such a purpose again. Why not? Because few real-world data analysis projects involve just one dichotomous independent variable and one normally distributed dependent variable. It almost seems like t-tests aren’t that important.

But they are, because they:

  • provide small-sample estimators; they don’t rely on asymptotic properties like many other statistical methods
  • illustrate null hypothesis testing in a simple manner
  • present the basics of frequentist statistics in the barest form possible
  • allow you to test the significance of regression coefficients, something much more common than two-group comparisons in day-to-day data analysis

Continue reading

psychology, that's random

Telling stories to the remembering self

I’ve been thinking about Kahneman’s remembering self, and how that part of the self needs memories woven into meaningful stories. Joseph Campbell’s monomyth structure offers a structure for telling stories to the remembering self. No matter how bad a particular experience is, you can probably make it into a story of struggle and growth, of confronting temptations and trials, of finding support where you thought you had none, of being in the wilderness then finally finding yourself again.

Here is Joseph Campbell on the temptations that a hero (or heroine) faces:

The crux of the curious difficulty lies in the fact that our conscious views of what life ought to be seldom correspond to what life really is. Generally we refuse to admit within ourselves, or within our friends, the fullness of that pushing, self-protective, malodorous, carnivorous, lecherous fever which is the very nature of the organic cell. Rather, we tend to perfume, whitewash, and reinterpret; meanwhile imagining that all the flies in the ointment, all the hairs in the soup, are the faults of some unpleasant someone else.

I do find it difficult that the actual experience of life is often so different than how it seems it should be. Telling stories about the bad experiences makes them make sense, turns the difficulty into something desirable, something that leads to learning and growth rather than something to be avoided or denied.

that's random

Detached, difficult, and boring

I’ve had two things open in my browser all day while I’ve been finishing off that policy paper: a review of The Age of Absurdity and Hugh Macleod’s ‘boring’ is underrated comic/post.

These seem to me to go together and also seem to be a message the universe keeps trying to morse-code to me. Gotta do the hard work, take it step by step, ignore the emotions, quit looking for excitement. You can’t succeed with that bursty stuff you used to do.

In The Age of Absurdity (as reviewed in The Guardian) author Michael Foley takes aim at the culture of instant-constant-increasing gratification and looks to the twin cures of detachment and difficulty, things I’ve been looking for myself over the past few weeks.


It’s not even as if we want what we have once we’ve got it. Foley calls this “the glamour of potential”, a relentless churning of desire by which the things we have are devalued by the things we want next. The only way out of the churn is “detachment”…

And difficulty:

The difficulty of change is aggravated in a society in which difficulty itself is avoided. Hence the study of science dwindles in universities (“Why submit to mathematical rigour when you can do a degree in surfing and beach management?”) and sales of oranges plummet because people will no longer take the trouble to peel them.

This avoidance of difficulty is what I’m looking for in the big data analysis project looming this week… I want to see if cultures higher on self-expression show lower math achievement, with the moderator of “liking math.”  In other words, do students who really enjoy math get directed towards high level achievement with it? Or, as happens in our culture (I think), do they get dissuaded because instead they might do something more authentic? I mean, what does it express about you if you like to do math? (Hint: you’re a nerd!)

In Macleod’s boring is underrated, I’m reminded that getting a book deal from your blog used to be big news, that making money blogging was hot, that blogging in 2004 was edgy and exploratory and experimental. It was exciting. Yes, yes, yes. I remember all that.

But I agree with him that boring <> bad (where, for you non-programmers “<>” means “not equal” or “isn’t equal to”). Blogging as a technology is boring but the ideas and connections and possibilities are not. Similarly, long hard work in support of an important goal may feel boring at times, but the results are not.

Still I can’t quite get worked up about detached, difficult, and boring. And I hope those words don’t start to apply to me just because I’m working on detachment, learning difficult topics, and slogging through the boringness.