Saturday, February 2, 2013

Silence and Flow - Minor updates to my preferred workstyle

Recently I found myself, half by intent and half by circumstance, cut off almost completely from the internet. No email, no twitter, no HackerNews. I don't have TV or cable. I also refused to answer my cellphone, handing it over to a friend to monitor. (He knew how to contact me in an emergency).

And.. I loved it. A few withdrawal symptoms for the first few days, but I've never been so productive in my life before.

 I'm thinking of making this kind of isolation + total focus a big part of work going forward. I'm thinking of going into "Alternate 4-6 weeks of focused work in isolation with 4-6 weeks of connected, 'normal' existence" mode (with few days of travel to random places in the middle).

Which might be really soon. I have a pile of mathematics texts to work through. My initial efforts to learn mathematics enabled me to break free of the 'take some random US business's database and build a web front end and call yourself a software engineer' phase of my career, and make a living writing interesting programs for interesting people doing interesting things. Now is a good time to ramp up on the math (and then write more interesting programs doing interesting things. ;-). Rinse. Repeat.).

Mostly my work (post the wasted enterprise dev years) involves a lot of programming and a little math. This year the ratio might be inverted. A scary, if exhilarating, thought - I find creating proofs way tougher than creating programs. With a good maths text, half a page is often a good day's progress. But on the other hand, it feels somehow more fundamental and satisfying.

One doesn't really need the internet to work through math texts. You don't even need a computer. You need a pile of paper, pens, and a good place where you can work undisturbed.

I did miss some of the news - I didn't get to know of Aaron Swartz's suicide for instance (poor fellow, R.I.P)  and the latest shenanigans of India's dysfunctional government, or the French invasion of Mali. But by and large, the world didn't change all that much in 6 weeks, and not having to follow or respond to  a constant stream of news and tweets and emails and phone calls is a refreshing experience.

The one complicating factor in all this is the prevalence of compelling MOOCs put out by Coursera, EDx et al. For 'attending' a MOOC, you need the internet, and then it is all too easy to drop into checking email, or peeping at Twitter and then you are back in the roar of the world flowing by. I don't quite have a solution yet. Working on it.

[1]  Phone calls are worse than any other communications medium because the damn thing actually *rings* and jerks you out of the zone to deal with (mostly) trivia. Switching into silent mode every time you want to do some work is a major pain, because (on my phone) you have to go through multiple button presses to accomplish the mode change. And then if you see a friend's missed call you feel guilty if you don't call back. But if you don't see anything, you can just get on with work and apologize later ;-).

Wednesday, November 28, 2012

The meaning of " I don't have a good (startup) idea"

I know a few dozen people who are looking for a 'good startup idea' and are seemingly held back from starting up only because they don't have an idea.

Which is curious, because there is a notion floating around that ideas are not important, or at least not very important wrt startups, and execution is all that matters. In other words you can have a so-so idea but execute brilliantly and succeed.

This is a slippery enough concept that nailing down exactly what is meant is difficult, leave alone arguing against it. But I've consistently encountered situations where talented, hardworking people say they have trouble coming up with startup ideas,

The other day I met a friend who I haven't seen for years, and since like everyone who is halfway technically competent, he is also "planning a startup". One of the questions he asked me was "So, do you have any decent ideas for a startup?" My answer "Sure, I have a few dozen"  seemed to surprise him. I suspect that 'I have plenty of ideas, sure' is a somewhat unusual answer amongst would be startup folks (who can also code). 

So this friend is a sharp dev, who can build anything he can conceive, has worked in startups, and still hesitates wrt starting his own company because he doesn't 'have a good idea'.


Yet another friend, ex dev, who works as a manager  in a (time and materials) software services company, tweeted the other day, about his (mild) frustration with an 'agile methodology' process bottleneck, and another friend who is an immensely successful entrepreneur, wrote back "Quit now. Life's too short for this nonsense. build something, give it your best shot and you'll love every minute of it.." And my manager friend replied  (I am not doing a startup because)"I don't have a good idea yet".

A third friend, very talented, with massive experience in business consulting, able to spot inefficiencies in half a dozen industries, still went into a loop of "What is a good idea for a startup?" and stayed there for a few weeks.

When you see something repeat three times, it is probably worth investigating what the underlying dynamic is.

Paul Graham in his recent essay "How to get Startup Ideas", says,

"The very best startup ideas tend to have three things in common: they're something the founders themselves want, that they themselves can build, and that few others realize are worth doing."

Turning these into questions, an alternative to asking yourself "What is a good idea I can build a startup on?" is to ask yourself

(a) What do you(personally) want (to exist in the world)?
(b) Can you build (the answer to the previous question)
(c) Do other people see/are other people working on the same opportunity?

These are easier questions to answer than 'What is a good idea for us to build a startup around?'. So how come people aren't attempting to answer them?

The rest of the essay deals with characteristics of good ideas vs mediocre ones, and how to select between multiple ideas and so on. (the whole essay is well worth reading).

Now let's look at someone who thinks very differently.


Paras Chopra, CEO of Wingify, is pulling in millions of dollars, from an office in India with a very small team, at an age where most of us are doing entry level jobs in IBM or Infosys or wherever. He found out the hard way that building 'cool' things don't necessarily bring in the money, so he decided to explicitly focus on making money. And he did (make tonnes of money).

Paras, like PG, is a doer, not a self help guru. I am fairly allergic to self help pablum by people whose job is selling self help pablum in the form of books or conferences (hence my dismissal of the whole "Lean Startup" (TM) idea, which is mostly tonnes of anecdotes wrapped around a sliver of a 'mother hood and apple pie" homily (hello 'agile'!). So yeah, I don't really believe in "set up a landing page and fool people into leaving their emails so  you can spam them" type approaches.

Paras wrote a trio of blog posts illuminating his philosophy --  "Sorry your "cool"  webapp is probably not going to make money", "How to find startup ideas that make money", and "Validate your startup idea by asking three simple questions"
Again his advice can be turned into the 'questions to ask yourself' format.

from his second essay, a more fruitful set of questions to ask yourself (than "What is a good startup idea?")

is

(a) What product is already making money for other people?
(b) Do you find this product (area) interesting /aligned with your skill set?
(c) What is a niche within the product area where you can launch a competing/disruptive product?


Again, these questions are more focussed, and easier to answer, than the overarching "What is a good idea for a startup"?

It is way easier to answer "What product is already making money for other people?" or "What do you(personally) want (to exist in the world)?" than "What is a startup idea"?

If so, and spending a day or two with these questions (and other 'how to do it' advice from people who have already walked this path) would generate half a dozen ideas, why do people still agonize over finding the right idea?

I think the real problem is more subtle. (Putting on my cynical hat) I think most people going around with "What is a good startup idea?" have no more intention of following through if they did get one, than do people who keep saying "I am going to write a novel (someday)" .  I don't think serious writers  (or even wannabe writers) greet each other with "So, do you have any nice ideas yet for a novel?". They might have rough  drafts of their next novel, but I doubt they spend time agonizing over 'good ideas' over beer.

Suppose you did get that 'perfect idea'. Are you really going to resign your job? And explain to your spouse, inlaws and kids, that you gave up a perfectly good job for an uncertain shot at changing the world? Then work insane hours and get into unknown territory like staffing, fund raising etc?

Isn't it easier to just talk about it? Say "I would be attempting a startup if only I had a good idea"?  The easiest way to never startup is to dismiss any embryonic idea with some form of "Yes, but ..."

My rather cynical conclusion is that a "good startup idea", for most people, is  something to think about occasionally, and talk about, and not really be executed upon. Which is perfectly fine, of course.

None of this is to argue that a startup is somehow a more noble endeavour than holding  down a BigCorp job, or consulting. I also suspect that startups are started either by people who've never held down a job (and don't want to), like the mythical Stanford students in their dorm, or by people who've had enough of the corporate bullshit that permeates most BigCo jobs, and decide to do a lifetime's work in n years and raise their 'Fuck You money', so they don't have to work anymore. People who are ok with their jobs don't (and probably shouldn't) try.

So why don't I do a startup? See above, tweak the details a bit.

All this applies to me just as to everyone else. My 'problem' with 'starting up' is not a lack of ideas. I don't have any trouble coming up with a few dozen ideas whenever I want to. I am a good enough programmer that I can build most things I can think of.

I don't want to start a company all by myself (this could change). Most people I'd like to work with live in the USA, and I have no plans to live there, ever. I also don't want to work on technically trivial projects (I'd be very unhappy working on  a Groupon clone, or doing Ruby on Rails consulting say). Meanwhile consulting on machine learning gives me my 'tackle hard problems' fix. What's not to like?

Are any of these insormountable obstacles? Not at all. I just choose to use these 'reasons' as excuses to go on doing what I do(vs actually living up to potential, what a scary thought ;)) .

 So I'm stuck, just like most other people. And I know why I'm stuck. I just try not to fool myself too much   ;)

Wednesday, October 24, 2012

On the "Do you want to be a programmer at fifty?" thing.

Once upon a time, James Hague asked an interesting  question on his blog.

"When I was still a professional programmer, my office-mate once asked out of the blue, "Do you really want to be doing this kind of work when you're fifty?"

James went on to identify two kinds of programming

Type A "work(ing) out the solutions to difficult problems. That takes careful thought, but it's the same kind of thought a novelist uses to organize a story or to write dialog that rings true. That kind of problem-solving is satisfying, even fun."


Type B "what most programming is about - trying to come up with a working solution in a problem domain that you don't fully understand and don't have time to understand... skimming great oceans of APIs that you could spend years studying and learning, but the market will have moved on by then ... reading between the lines of documentation and guessing at how edge cases are handled and whether or not your assumptions will still hold true two months or two years from now.. the constant evolutionary changes that occur in the language definition, the compiler, the libraries, the application framework, and the underlying operating system, that all snowball together and keep you in maintenance mode instead of making real improvements."

He went on to state that while he'll continue doing Type A programming, he isn't particularly interested in Type B, (presumably at fifty).

I was looking forward to some good discussion on this, but HackerNews (which, in spite of its flaws still has no competition) went off into some  tangents primarily about ageism in the software industry and there was surprisingly little discussion about what James actually said.

Now, is ageism a problem? Yes, it is. As people grow older, they are expected to do anything but programming. It is a cultural thing and not necessarily logical. I know someone who is a good programmer, but left Bangalore for a decade (programming all the while) and now can't get an interview (let alone a job)because "oh you have 18 years experience, we are looking for people with two years of experience.Sorry".

 So, yes ageism is a problem, even in Outsourcing Land, and there is plenty to be discussed, and action to be taken, with respect to ageism.But that is a topic for another day and  isn't quite the problem addressed by James Hague in his blog post.

In this post, I'll try to explain what I think (and it is just that, my opinion ymmv etc etc) about "Do you really want to be doing this at fifty?"

The essence of the question is "Do you want to be doing this(at  a future time point)". The question addresses the (evolution of) motivation to program, and James goes on to state that his motivation to do a  certain type of programming (unfortunately this is the more dominant type  of programming worldwide)  decreases with increasing age.

The question of motivation with respect to career activities has been discussed by a wide variety of people and a lot of research has been conducted. One  interesting insight has been articulated by Dan Pink - in his book "Drive - The Surprising Truth About What Motivates Us" he identifies 3 factors  that motivate us  (or demotivate us) to undertake and pursue any activity.

(1) Mastery - getting better at what you are doing 

(2) Autonomy - the degree to which you can direct your activity.

(3) Purpose (or meaning) - doing something that really matters.

If you get high scores in any of the above, ideally all three, and more importantly get more and more of all the above as your programming career progresses, of *course* you'll be programming at 50. Or 60. Why wouldn't you?

The problem of course, is that in most programming jobs you either hit a declining slope or at best, plateau, with respect to one or more of the above as you age. If you are on a team of 50 people, maintaining some legacy leasing system written in Java, with business analysts doing the business thinking and you are converting their thoughts into code you are being a scribe for other people's ideas in a rigid and ageing language, in a context where you are an expensive 'resource'.

In general , even at many product (vs services) companies,  a 'line programmer' has low levels of autonomy - other people - product managers, business analysts etc etc - tell him what to do. Legacy codebases constrain technology choices. His 'mastery', while not non existent, is of a shallow and frothy kind (hey I use Rails today instead of J2EE yesterday! Node.js vs rails blah) and writing the n-th business app pulling data off a database and putting it on a web page for some corporate drone to use to update his TPS reports crushes 'being part of a higher purpose'. Little autonomy, modest mastery, non existent  purpose. No wonder few people  want to be doing this kind of programming at fifty.

Thankfully, other kinds of programming do exist. John Carmack of Id Software is still programming in his forties because programming (and till recently, being a majority shareholder of a cutting edge games company!) helps him in maximizing all three  attributes.

Programming is a skill, like writing. Unlike writing, we live in a society where most people are code illiterate. And coding ability has (some) economic value. "Software is eating the world" etc, and so anyone who is comfortable with coding can exchange that skill for money. The deeper question is whether you can trade increasing experience in the skill of programming for increasing amounts of money (and mastery and autonomy and purpose) as time passes. For most people that function plateaus and then stays steady or declines.

If you were someone who knew how to write, but lived in an illiterate society you could exchange that skill for money, by being a scribe at so many cents per word. You write people's letters and wills etc and you get paid for it by word count. But if you did it for thirty years, and you are still writing letters for people when you are fifty, would you be satisified with your career? What about when your customers move to that desperate youngster who offers a lesser rate per written word?

A novelist uses writing in a different manner than someone who sets up as a letter writer for illiterate people. A novelist is trying to do something that uses writing grammatically correct sentences as a base skill, but the core of his work, plotting, characterization, dialogue, world building, etc lie on a plane well above deciding whether to put an i  before an e, or vice versa. And you don't even need much base skill.  Many people are pretty bad at grammar and still write best selling or world changing books.

Generalizing, the (conceptually) shortest step to getting away from the 'path to ageist irrelevancy' for programmers is to find a way to make money by transcribing your own ideas to  code. This might involve, for example,  stepping away from time and materials services  types of programming to product development. If not by yourself, then as a part of a small team. Even if you are still technically an employee, your are much more autonomous in small teams and companies (and codebases).

A second way out of a programming career deadpool track is to move to something related where programming skill actually helps in a major way, but it isn't the core of your job.

 If you are a Computer Science researcher who is also an excellent programmer, your primary job is the creation of new knowledge (aka research, embodied in published papers) but your programming skills will help.

If you are a (technical) startup founder using cutting edge languages and algorithms to build a superior product, your primary job is to satisfy users and pull your company ahead of the competition, then superior programming ability can help.

If you are a finance expert who can also code, you probably have a significant edge over your competitors who have to depend on the software people to come in after the weekend to prototype your idea.

Programming skill amplifies effectiveness in almost everything you can do.

Of course you could find yourself in the same situation in your new career. If you still lack money, autonomy, mastery and purpose, you are back at square one.  That said, being "an excellent programmer and a good  X" seems like a decent plan.

The idea that a programmer always has to work in a half understood domain transforming some one else's ideas into code is just that, an idea. It is a dominant idea, but nothing really stops anyone from mastering an interesting domain or acquiring a complementary  skill  in addition to programming.

That gets me to what I think is the right way to go about 'career planning'.

Decide what increased levels of autonomy, mastery and purpose mean to you.  Figure out what you need to do to get to that point. Then do whatever it takes.

If increased programming skill will move you towards increasing one or more of the three attributes, work on it. If something else (like writing skill, or knowing a domain, or getting good at sales, or going to medical school) looks more promising, work on that.  Assembly  line programming inside 'the industry', converting other people's thoughts into code in stone age languages, is a beginning. It need not  be the end.

To conclude, will I be programming  at fifty? I think so (these days I do as much maths and stats as programming, and everything feeds very nicely into everything else), but at fifty,  I'll be writing novels, not scribing letters.

Sunday, October 14, 2012

Why I (need to) write (more)

For some context on what sparked this, see this excellent post and discussion on Hacker News, (for those so inclined)

The reason I used to blog somewhat regularly is very simple - I found that I could take some half formed thought and flesh it out by writing about it.  Sometimes this would result in something actually useful to  people. Most often, I would just end up 'emptying my head' and think new thoughts.

And then something happened. I got an audience. And comments. And emails.  And controversies, and flames.

All good, because I didn't (and don't) *mostly* care what people thought of anything I wrote (or what people think of me for that matter) .  But there *is* a small element of reactivity and  friction when you know persons X, Y and Z will be reading what you write.

"Hmm is this too harshly worded for friend X. After all he is a big Ruby fan and if I say Ruby is a particularly  brain dead language, would this ruin his day?". And then I have to write stuff twice. First write down what I really want to write, and then go through it with and delete stuff or add more explanatory stuff and cautionary qualifiers and so forth.

I am not the only person facing this. People who are much better at writing than me apparently face this too.

Someone asked Paul Graham (he of the glowing essays fame) on Hacker News

"what's it like to have your every written (or spoken!) word analyzed by a bunch of people? Esp. people that you end up having some form of contact with.

It seems like it would be difficult to just have a public conversation about a topic. Do you think about that much when you write?"

and PG replied to say

"It's pretty grim. I think that's one of the reasons I write fewer essays now.
After I wrote this one, I had to go back and armor it by pre-empting anything I could imagine anyone willfully misunderstanding to use as a weapon in comment threads. The whole of footnote 1 is such armor for example. I essentially anticipated all the "No, what I said was" type comments I'd have had to make on HN and just included them in the essay."

If pg can't escape this fate, I sure can't.

but otoh I am less concerned than pg about whether someone mis-understands me etc because, I am (comparitively) not famous,  and I am not writing essays just spewing out (comparitively) unpolished *blog posts*. I could deal with misunderstandings just fine.


What I found harder to deal with was -- Twitter.

Once I started tweeting regularly, I found I could just try to distill whatever I was thinking about and just tweet about it. 140 characters is pretty good as a constraint. And since I regularly purge my twitter following of idots and nutcases, I am fairly sure I can convey exactly what I want to, and most people following me would understand (and if not clarifications are just  1 < n <   3  * 140 character tweets away.


But as good as twitter is, 140 character tweets aren't as good as multi paragraph blog posts to *explore ideas* (vs expressing their seed forms concisely). My writing has suffered, though like riding a bicycle it should come back pretty fast, and then I'll go around improving it.


The goal of my writing remains the same. I write to explore thoughts and ideas and 'empty me head'. No more. No less. 

One thing I am doing differently this time is to pay even less attention than usual to comments and reactions, and not bother clarifying what precise shade of  meaning I intended to convey and so forth. This is just me writing a letter  to a friend every  other week or so. It just takes the form of a blog post other people can read.




And so here goes.

I'm back.



Monday, September 24, 2012

Renewal. Learning. Stuff

I haven't written anything here for more than a year.

 That will now change.

 I was too busy and twitter was interesting enough - 140 character limits are good training for conciseness - that I didn't miss blogging all that much.

But writing longer pieces have their own advantages, and I hope to write at least one entry a week for the next year or so.

 Stay tuned.

PS: I hate the changes Google has brought to blogger.  A change of platform is on the agenda. I just don't  have the time right now. 

Saturday, October 8, 2011

An unimportant person''s comment on Steve Jobs's death

Context: Everyone and his dog is hyperventilating on the internet about the death of Steve Jobs.

Here is my opinion (which like most opinions isn't worth very much, but hey this is my blog).

All men are mortal.

Steve Jobs was a man. (A great man, but still, a man.)

[Modus Ponens] Steve Jobs was mortal too.

Now he has died. The world endures. Life goes on.

Your (and my) time to depart will soon be here. The world will still endure. And life will still go on.

I read somewhere that the one regret most people have at the moment of death is about how they should have done X or Y instead of A or B.

Get back to work. Do X or Y instead of A or B. Die happy, when your time comes.

To the degree one admires Jobs, emulating his virtues in your life is a more fitting tribute than another silly comment about how he was as influential as Plato and Aristotle (an idiot actually said this on HN).

Update: on Stallman's comment on Steve Jobs' death. People have different ideas on whether a person's achievements were good or bad. This affects their judgement of whether a person's death was "good for the world" or not.

Stallman thinks that the end of Steve's influence on computing (note: he clearly distinguished it from Steve's death itself) is a good thing. And said as much.

I don't agree.

I think, for all his faults (and like you, me and every human who ever lived, he had some) Steve's influence was beneficial (overall) and I wish he'd lived longer.

But I also think it is ok for people (including Stallman) to express their opinions, even when I find them not in agreement with my own.

I look forward to the day when everyone (including me) will shut up about how other people should think exactly like everyone else.(or else we'll all get all self righteous and puffed up and hyperventilative).


And now I'll go back to coding. (Thank You for reading this far.You really should be doing something useful instead!)

Monday, August 15, 2011

On Owning a Kindle



I was gifted a Kindle a month or so ago. I like it for what it does.

Should you buy one? If you are a book lover, one of those people who always have a book on hand, or reach for one when you have an hour to spare, you definitely should. If you read mostly technical or math books (which require a lot of flipping back and forth and good rendering of code or equations)or research papers, you won't get as much benefit as you ought to.

I am well satisfied with the Kindle for allowing me to lug around about 300 books (I still have almost 3 GB left) so I can read on the bus, while waiting for someone, etc. I would have loved it if I could read math books and papers (pdf rendering on the (small) kindle is terrible) and also scribble notes (the kindle "make notes" functionality is awkward and unusable) but e-ink based readers are still in their early days. For what it does (enable you to carry around a few hundred fiction books it is awesome. For example, I have all the 20 Aubrey Maturin books and the dozen or so Jim Butcher books in a 6 inch device. (E-Paper blows away IPad's screen for reading.)

If I'd received the Kindle before the release of George Martin's utterly terrible "A Dance With Dragons" (I should write a blog entry one of these days on how terrible it is - suffice to say that the man has lost his touch) I could have spent 11$ on a kindle version instead of 54 $ (18 for the book, the rest for postage to India). The Kindle shines for fiction and light non fiction books. And you can avoid paying for the kindle editions by downloading "pirated" versions if you know where to look. I suspect it would work well for magazine subscriptions too ( at least for those in which the written word is more important than glossy pictures).

Somewhat tangentially, someone should write a piece of software that works like LaTex for math but generates flowable text. Tex is (print) page oriented.If you could just take a LaTex file and generate a kindle readable document out of it,I suspect a lot of math/tech papers would find their way on to e-readers very fast.

After having used the kindle for a while I am not surprised that Amazon sells more e-books now than paper books. I suspect the Kindle is a very potent weapon in Amazon's arsenal, that its competitors underestimate. If they make it work in the Indian context, (Amazon plans to launch in India in 2012 - I have no idea how much of a role they plan for the Kindle here), their competitors will get swatted aside like so many flies. (Hmmm I should write a post on how I see the Amazon-Flipkart battle shaping up in India. Interesting times we live in).

Meanwhile, if you are a reader and can afford to buy a Kindle, you should. It (or something like it) is the future of reading.