## Learning from mistakes

[Inspired by Dennett’s Intuition Pumps]

My current approach and where I’d like to be works 1:1 or in a small group, much more difficult to do in a lecture-style setting.

When I’m trying to teach a new concept – programming, algorithm, mathematics – I try to give students a bit of structure and context and get them to develop the solution. The amount of structure and context varies from student to student, and its not easy getting it right. Done well, students reach that aha! moment on their own. I’ve been doing this unconsciously for quite a while, but its probably time to do this with more structure (hah!).

One more aspect I would like to add is having students reflect on failed approaches. This may also require reducing the context and allowing them to try out a variety of approaches.  A lot of learning can happen in this analysis. The downside is that it needs time and it needs willingness to expend the effort.

I’ll eventually organize my thoughts, but a few reactions:

• This book should be dedicated to Steven Pinker. No matter what aspect of business one may pick up, here’s how to look at it in a positive light
• The amount of data to back up the positivity is of course, amazing
• I’m going to skip the ‘what-about-ism’s, but that is pretty much what was going through my head as I was reading this book

This book is necessary – (big) business has been vilified for too long, and the default is to assume that it is evil. This book pushes the pendulum towards the center and swings it all the way to the other end!

Transcript

## Background

This is the first article I recommend everyone read (currently, ‘everyone’ refers to students or recent graduates). Every paragraph is full of insights. It encourages a proactive approach to the world around us. It lays out an approximate framework for anyone interested in thinking about what they are doing, and where they would want to be.

The following notes are a summary of my observations and formulations from the article. These are a combination of the obvious (“he says it right in the paper!”) and bringing together concepts scattered across. It is a work in progress!

## Why do great research work

Unlike the paper, I believe each of us has to define what ‘great’ is. We should be doing work that matters to us, and this needs to lead to significant scientific breakthroughs. Also, the insights apply equally to work (as opposed to research)

## The (mis-)characteristics of a great scientist

What is common among the great scientists? The lazy narrative is that they achieved greatness because of:

• Luck: Somethings are happenstance, but for the most part, make your own luck!
• Fortune favors the prepared mind – Louis Pasteur
• If others would think as hard as I did, then they would get
similar results – Issac Newton
• Brains: Most likely we have enough for what we set out to do. It just needs unlocking.
• Age: Depends on what we are doing. In some cases, the audacity of youth is needed to challenge existing dogma, on others, the wisdom of maturity is required to tease out insights.
• Drive: This is so important, I discuss it separately below.
• Ambiguity: College exams are typically re-statements of problems already seen. However, the real world is rarely straightforward.
• Commitment: this is necessary to get the subconscious to work for us. And this leads to that nebulous factor called creativity.

## Drive

There are so many things to unpack from this one paragraph! In fact, this was the one thing that I remembered from the first reading of the paper. Here’s how I visualize one part:

Knowledge and productivity are like compound interest. …. The more you know, the more you learn; the more you learn, the more you can do; the more you can do, the more the opportunity.

Hamming anticipates Gladwell’s 10,000-hour rule, noting the importance of effort intelligently applied.

## Ego and Perspective

I recall colleagues who insisted on doing things their way, no matter the effort and energy it would consume. Much better to figure out what is important, which fights you want to fight and drop the rest.

## Be proactive

Hamming does not explicitly say this, but throughout the talk, we find examples of him identifying areas of improvement and actively working on these. Examples are:

• He has lunch with other teams and has an agenda of what he wants to get out of these interactions.
• He consciously sets aside time for ‘great thoughts’ (in a sense anticipating Google’s 10% approach).
• He starts out with being nervous giving a speech, so volunteers to do more of them!
• Opportunity

## The ‘Big Picture’ view

Closely related to the above point is Hamming’s consistent approach to thinking about the underlying structure of the reality that he observes (the entire talk is an example of this, of course). Why do some scientists do great work? What is the larger problem that I am trying to solve? How can we generalize this work? Is it better to do x instead of y. Why? (an example: work with the door closed or open?). Who is more effective in a meeting? Why?

Too many times we just accept things as they are. It’s a seemingly small step to question things, but the consequences can be huge.

## The interplay of factors contributing to great work

Great Work $=\left(d\times c\right)^a$
Where

$d =$ drive
$c =$ commitment
$a =$ ability

The factors that are in our control — drive and commitment are the base of the exponent. The more we invest, the greater the returns. Our ability is the exponent, and this can grow if we put in the effort.

## This Happened…

Eliud Kipchoge ran a sub-2 hour marathon

Transcript

# Background

This is the first article I recommend everyone read (currently, ‘everyone’ refers to students or recent graduates). Every paragraph is full of insights. It encourages a proactive approach to the world around us. It lays out an approximate framework for anyone interested in thinking about what they are doing, and where they would want to be.

The following notes are a summary of my observations and formulations from the article. These are a combination of the obvious (“he says it right in the paper!”) and bringing together concepts scattered across. It is a work in progress!

## Why do great research work

Unlike the paper, I believe each of us has to define what ‘great’ is. We should be doing work that matters to us, and this needs to lead to significant scientific breakthroughs. Also, the insights apply equally to work (as opposed to research)

## The (mis-)characteristics of a great scientist

What is common among the great scientists? The lazy narrative is that they achieved
greatness because of:

• Luck: Something are happenstance, but for the most part, make
• Fortune favors the prepared mind – Louis Pasteur
• If others would think as hard as I did, then they would get
similar results – Issac Newton
• Brains: Most likely we have enough for what we set out to do. It
just needs unlocking
• Age: Depends on what we are doing. In some cases, the audacity
of youth is needed to challenge existing dogma, on others, the wisdom of
maturity is required to tease out insights
• Drive: This is so important, I discuss it separately below
• Ambiguity: College exams are typically re-statements of problems
already seen. However, the real world is rarely straightforward.
• Commitment: this is necessary to get the subconscious to work
for us. And this leads to that nebulous factor called creativity.

## Drive

There are so many things to unpack from this one paragraph! In fact, this was the one
thing that I remembered from the first reading of the paper. Here’s how I
visualize one part:

Knowledge and productivity are like compound interest. …. The more you know, the more you learn; the more you learn, the more you can do; the more you can do, the more the opportunity. Hamming anticipates Gladwell’s 10,000-hour rule, noting the importance of effort intelligently applied.

## Ego and Perspective

I recall colleagues who insisted on doing things their way, no matter the effort and energy it would consume. Much better to figure out what is important, which fights you want to fight and drop the rest.

## Be proactive

Hamming does not explicitly say this, but throughout the talk, we find examples of him identifying areas of improvement and actively working on these. Examples are:

• He has lunch with other teams and has an agenda of what he wants
to get out of these interactions
• He consciously sets aside time for ‘great thoughts’ (in a sense
• He starts out with being nervous giving a speech, so volunteers
to do more of them!
• Opportunity

## The ‘Big Picture’ view

Closely related to the above point is Hamming’s consistent approach to thinking about the underlying structure of the reality that he observes (the entire talk is an example of this, of course). Why do some scientists do great work? What is the larger problem that I am trying to solve? How can we generalize this work? Is it better to do x instead of y. Why? (an example: work with the door closed or open?). Who is more effective in a meeting? Why?

Too many times we just accept things as they are. It’s a seemingly small step to question things, but the consequences can be huge.

## The interplay of factors contributing to great work

Great Work $= \left(d\times c\right)^a$
Where d = drive
c = commitment
a = ability

The factors that are in our control — drive and commitment are the base of the exponent. The more we invest, the greater the returns. Our ability is the exponent, and this can grow if we put in the effort.

## Bringing Sanity to the Campus Recruitment process

For the last 2 months, my interns have
been MIA, because “companies are coming to campus.” When I speak with students
about the process that they (the companies) follow and the torture that they
(the students) undergo, a few things become apparent:

1. Every company does things its own way. In broad terms, the process is similar, but the details are different enough to make each one unique. To an outsider, these processes seem to be ad-hoc. None of them are like what Shell does (incidentally, Shell recruitment for an IT position also doesn’t seem to follow the normal evaluation process!). Consequently, replacing one process with another won’t make much of a difference in terms of outcomes (see the next points)!
2. I’m curious if any company examines the effectiveness of its selection process. Wouldn’t it be great if they kept the evaluations (and hence expectations) of candidates, and then revisit the data in a year or two to figure out how accurate they were? And then tune the process to improve the outcomes for the company. Google did this a few years back with not-so-surprising results (link).
3. Related to the above, what is the cost-benefit tradeoff of this exercise?
4. In all fairness to the interviewing process, it is difficult to judge the suitability of each candidate in a packed 1-day schedule. But still…. a few of my students would be fantastic additions to any team, and they’ve been passed over.

From the student’s perspective, they pretty much lose out on two months of time that could be spent on more productive endeavors. And don’t even get me started on the stress that they put themselves through as they do not get selected with each successive company.

A serious drawback of the
current system is the power difference between companies and students.
Companies fight for the privilege of being first on campus. They can accept or
reject students from the pool that applies. Once a student has been accepted,
they can no longer apply to other companies. They may have preferred a company
that comes to campus later, but didn’t want to risk not applying for the first
company! So students end up interviewing for companies in the order that the
companies are scheduled, not in the order that they prefer!

To
summarize, the current process does not seem efficient to the company and even
less so for students. Here is a modest proposal based on the observation that
most positions at the entry-level fall into a few categories (developer, IT
analyst, customer support, …) and technologies (cloud, databases, Java,
Android, …). From the cross-product of role and technologies above,
administer one set of evaluations that cover aptitude, technical
knowledge, specialized skills (e.g. implementation) and an interview. Companies
can use these results and calculate their own weighted scores to rank students.
Students also rank companies based on their criteria — job profile, company
profile, salary, etc. You can guess where this is going: this is an instance of
stable matching and
we can use some variant of Gale-Shapely to generate an optimal matching
(whether this is optimal for the student or the company is left as an exercise

Students/companies end up with the best
possible outcome. A student gets the best job that they can get from the list
requirements who want to work for them. The process takes a week at the most.
Students don’t waste time and are relieved of unnecessary stress. Collectively,
companies save a huge amount of time as well. The data generated can be used to
track and modify/improve the process as needed.

This can be done at one institute, but
really speaking, should be extended nationwide. It is not feasible for every
company to visit every college they would like to go to. A process such as the
above vastly increases the pool of good candidates and minimal cost.

The main objection that
companies may have relates to the validity of a common system. (aside: in an
ideal world, interviews should be unnecessary and college grades should be
enough to evaluate a graduate hire. The fact that interviews exist indicate that
something is amiss in the grading system. Funnily enough, companies use GPA as
a cut-off to filter candidates, but not as the only deciding factor). However,
the current interviewing system isn’t an efficient system anyway and I believe
that my proposal will be no worse than what companies do; and will be much
better when the cost is taken into account.

A second objection could be the loss of
prestige points for no longer being a “day-zero” company. With this system, if
you’re getting the best candidates, the ego is the only thing that is bruised.

The
nature of recruitment is going to change in the near future. Even now, I
believe TCS no longer goes to campus, they have an online test that anyone can
give, and candidates get selected for following stages based on the performance
on this test. All of a sudden, a student from a remote college that TCS would
not have visited has the opportunity to apply.

It’s not just recruitment that is
changing. Don’t look now, but the nature of “work” and education itself is
undergoing a transition, which is a topic for another article.

We’re in for interesting times…

Search