Innovation Is Not Shovel Ready Stuff

Posted on by David Wright

Here’s Mandel linked to from his blog:

We have only two ways out of our current global economic mess: innovation and inflation. And as the saying goes, we should hope for the best (more innovation) and prepare for the worst (higher inflation).

I agree with this simply because I try to read a lot on this topic and I haven’t come across a third way.

The problem is that innovation / productivity growth isn’t obviously happening anywhere. The bigger problem is that there is a lag between when innovation ‘happens’ and when it results in real economic growth.

As others have pointed out, our economy is increasingly dominated by services, simply because we’ve gotten pretty good at manufacturing stuff efficiently. Forget about the China crap, that’s a red herring. We were going to hit this wall eventually (think about 3d printing as eventually displacing even the cheapest Chinese labor).

And the thing with services is that, in an information economy, we get people handling and processing lots of data. And people are bad at handling and processing data. They make datapiles not datasets.

The problem with training people to be better at using machines is that you either need to teach old dogs new tricks (how many CEOs are this enlightened?)  or you need to wait until the slow march of demographics increases the IT IQ of corner offices around the world.

The point is this: if innovation needs to reform the services sector, the people in that sector need to become better programmers. And that ain’t happening soon.

Pro Tip

Posted on by David Wright

If you’re an eater of all-natural penut butter, like I am, one of the most infuriating things in the world is having to remix the oil and solid paste that gravity pulls apart after the jar’s been sitting on the shelf for a while.

When I was a kid, I used to pull out a knife or spoon or something, shove it into the jar and churn. It spills over the top, I get all greasy… What a mess.

Well, folks, I don’t do that anymore.

NOW, I take the jar and simply plop it into the cupboard upside down and wait a day or so.

What happens? Well that infuriating separation process is thrown into reverse and gravity’s my friend now. When I pull it out again a little while later, I get smooth and consistent creamy goodness.

This isn’t perfect, of course, as there might be the odd clump hidden in there somewhere, but by the time I find it, the jar is low enough that I can dispose of the offending sludge with a quick flick of my knife.

Use this knowledge wisely, readers.

You Are What You Do All Day, Until You Are

Posted on by David Wright

Malcom Gladwell has a knack for catching the public consciousness, doesn’t he?

Ever since that Outliers book (I’ve never read a Gladwell book), much of the really interesting writing I’ve come across probably owes its inspiration to him. Every personal experiment runs into its Gladwell moment:

Let me get to the big point here: You are what you do all day.

What does all this research about your body adapting to circumstances tell us?You are what you do all day.

What does all this research about brain plasticity and rewiring tell us? You are what you do all day.

That probably scares the shit out of a lot of people. And it should.

What did you do today? Is that what you want to be a genius at in 13 years? Is that what you want to become?

Nietzsche spoke of the eternal recurrence. Basically, you should live in such a way so that if I said to you, “You’re going to have to repeat this life over and over again for infinity” that your response would be “Sounds great.”

Do you feel that way? If not, are you going to change what you do tomorrow?

Your brain will change to adapt. Your body will change to adapt.

But will you?

And I’m sitting at one right now, actually. These Stanford courses have whetted my appetite for serious learning and I’m completely uninterested in actually shelling out real dough fo the privilege of packing my brain.

So I’m going to do the reverse of making my hobby my day job. I’m going to, once again, make my day job my hobby. I’m going to become an actuary.

By a Gladwellian definition, I’m already an actuary, or getting close, anyway. It’s what I do all day and, though I don’t quite have 10,000 hours under my belt, I’ve probably got about 3-4,000 hours of pure actuarial experience in the bag. My job has been migrating this way over the last five years or so and when I moved down here to New York it’s pretty much become full time.

And I love it. I really do.

My TIPPING POINT came when I downloaded an actuarial text for pure extra reading and reference for work. I thought to myself: how many people are stupid enough to read something like this and choose to spend zero extra effort to join one of the highest-status clubs on earth with infinite job security and economic prospects?

I had a flick through the syllabi of the 5 exams I’m going to need (I did one years ago between CFA exams) and I was surprised at how much I already know. See paragraph above. Facepalm.

So the first one is in January and I’m going to pack as many as I can in next year, which is probably going to be three. If it’s going super-duper well, I might chance a fourth. We shall see.

Machine Learning Course Notes

Posted on by David Wright

Still at it. I am loving this course.

Today we go to logistic regression, which is a fancy term and means that it is used to predict binary outcomes.

Binary outcomes are super-risky evaluations because while math doesn’t like discrete data, humans love it. Think about medical evaluations: you’re either ‘sick’ or ‘not sick’ in your own mind, but according to mathematized science, you have a particular combination of abnormal scores on a blood test, etc. These combine to produce a binary evaluation, “sick”, but that’s only because we need to cross a decision boundary to take action (begin treatment).

Logistic regression tackles this in a few ways. First, it lets you set where you think your decision boundary is going to be, when evaluated against a series of inputs (blood cell count, let’s say) and set an overall threshold for the evaluation. Let’s say that you assign a certain number of points to each input: 50 points per 100 red blood cells, -20 points if you work out every week, + 10 points for every cigarette you smoke. Then we say, if this person has more than 750 points, we declare them sick.

Now this point system isn’t perfect, there will be people we should have labeled sick with 300 points and people who are actually fine at 1000 points. Logistic regression gets around this by imposing a non-linear cost for being wrong. When fitting the curve (and figuring out that 750 level), the algorithm is penalized more heavily for misses at 1000 points than at 500.

Error in logistic regression is ALWAYS non-zero.

People Are Terrible With Counter-Factuals

Posted on by David Wright

Here’s an interesting piece: “10 Years Into the iPod Revolution”. I tend to get really irritated with this kind of attribution. My instinct here is to say: it would have happened anyway.

They dig up an interesting review of the original iPod:

People used to argue whether the trend was toward an all-in-one gadget that does everything as opposed to a collection of specialized gadgets. If I’m right about the iPod, both sides of this argument are correct; people will use one comprehensive iPod-like storage and connectivity unit in combination with every specialized peripheral you can think of. As before, something designed for digital music will spread across other areas of technology. Descendants of the iPod MP3 player will replace the PC as the hub of your digital life.

You could look at that last sentence and say: “OMG, he gets it. Apple was destined to make the ipad”. But you’d be skipping over some pretty important information.

First, the ipod’s descendents have hardly become the hub of anything. iCloud is making a play for this, but only within the Apple walled garden. We shall see whether this works.

For another, the iPod was simply the best HIGH-END mp3 player out there. There was always going to be a high-end mp3 player and Apple just crushed that market. Without them, there would have been another and maybe we’d be talking about that one instead.

My first iPod was the shuffle, which was, as far as I can tell, the first real mainstream product Apple ever made. Then Apple found its home in the cell phone market and its exploitation of gigantic personal discount rates. Presto: expensive products seem cheap.

Convergence between mp3 players and cell phones was always inevitable. Apple was the exception, I think, in that no other mp3 player manufacturer made the leap to phones. In every other case, the leap was for phone makers to just add mp3 functionality.

I don’t want any of this to suggest that Apple’s innovation machine wasn’t (isn’t) awesome. That’d be stupid. But to say that they’re more than, say, 10% better than the next rival is overdoing it.

Today’s worst mp3 players are a thousand times better than the original ipod. Apple’s cleverness buys it a bit of time, but that’s all.

Some Links

Posted on by David Wright

Been light on posting as I dash myself against the rocks of Machine Learning programming. Here are some neat things I’ve read in the last 20 minutes:

1. What is Dark Matter?. No answers, but an illustration of the infurating paradox: there’s almost unquestionably a THERE there, but nobody has any idea what’s there. The most important problem today in basic science?

2. Steve Jobs’ best interview ever? Stay tuned.

3. The Post-Industrial Economy. Ridiculous phrase for a concept I really care about. Bottom line: people will have ever greater power to make their own stuff. 3D printing, programming skills, ridiculously cheap education on just about anything. Try to wade through the consultant-jargon and think about these concepts.

My blind spot for human progress is materials science and 3d printing. I know nothing about these two massively important topics so they’re not on this list. If I could stitch together even one coherent sentence on the matter, I’d be all over it.

Machine Learning Course Notes – Bittersweet

Posted on by David Wright

Finished this week’s exercises in a 5-hour marathon starting at 4:30am this morning. Today’s meta-lesson: implementation is way harder than reading slides and ‘kinda getting it’. My god is it hard to actually write a program that uses even what appear to be simple concepts.

So there are three tracks for this course: first is the spectator track (my term), where you just do the basic assignments (enough to be dangerous and spew plausible-sounding BS).

There’s the ‘advanced’ track, which I’ve chosen, which asks you to do some actual programming assignments (this morning’s marathon). Within the advanced track there are ‘extra credit’ assignments, which ask you to implement even more of the course material in Octave (a programming language). I haven’t gotten to the extra credit stuff. More on this later.

The final track is the ‘real’ track, where you pay real money, show up to class and all the rest. I read a discussion thread on the course website that speculates that my ‘advanced’ track covers about 40%-50% of the real course material. The real course is about 1.5x as long (3 months instead of 2), so let’s say we’re about 60%-75% of the pace of a real university course.

I’m starting to think it was a mistake to take two of these courses. I just don’t have enough time to learn everything I want to learn. I want to do the extra credit stuff, because what’s the point of reading the slides on stuff if you don’t REALLY get it? And my first crack at the extra credit stuff shows that I don’t REALLY get it.

And there are all these dudes (yes, all dudes) carpet-bombing the discussion boards who obviously REALLY get this stuff, while I only kinda get it. How many times in University did I wish I were smarter?  That I wish I had really learned the background material in high school like I should have and I could have picked this up quicker?

Anyway, I’m done complaining and it’s just too time-costly for me to learn more of this right now, so I won’t. I wish it were different but that’s just too bad for me, isn’t it.

Machine Learning Course Notes

Posted on by David Wright

Not much to report or record at the moment, as all of the lectures and exercises this week have the goal only of teaching us the Octave programming language, which is an open source version of Matlab.

I’m constantly impressed by the power of expressing spreadsheets as matrices and vectors and thinking of analytical operations as expressed by linear algebra. Lots of new things to think about here, but gotta work through the programming exercises first.

Trends in Computing

Posted on by David Wright

Here is Celent on what the trends are that affect technology in insurance companies:

  • Data as platform
  • Analytics
  • Cloud computing
  • The move from server-centric architecture to service-centric architecture
  • IT security
  • Data privacy
  • Social platforms
  • User experience

Pretty boring list. And, I think, a list that doesn’t cut to the chase. Here’s my list (in order from least to most ‘social’ in nature):

  • Speed (Moore’s Law)
  • Memory: that’s disk capacity (and speed), and working memory capacity
  • Bandwidth
  • Development of user-friendly interpreted languages
  • Standards integration

To me, everything else comes from one of these trends.

Want to know what’s going to happen tomorrow in technology? Take one or more of those trends and imagine what happens when you double its current state of the art.

We Need More APIs

Posted on by David Wright

THE most constant pain in my ass is getting equivalent datasets in radically different formats. Every single insurance company records the same friggen data in different ways.

And if that wasn’t enough, whoever downloads this data then gets their grubby little paws all over it (usually mannually effing around with rows and columns in excel… [vomit]) before sending it to me.

I want machine data, I do NOT want people data. It is infuriatingly complicated cleaning up these datapiles and would all be better if they’d just give me access to their machines.

I’m reminded of the Matrix.