A Glimpse of Google, NASA & Peter Norvig + The Restaurant at the End of the Universe

I came across an interesting talk by Google’s Peter Norvig at NASA.

Of course, you should listen to the talk – let me blog about a couple of points that are of interest to me:

Algorithms that get better with Data

Peter had two good points:


  • Algorithms behave differently as they churn thru more data. For example in the figure, the Blue algorithm was better with a million training dataset. If one had stopped at that scale, one would be tempted to optimize that algorithm for better performance
  • But as the scale increased, the purple algorithm started showing promise – in fact the blue one starts deteriorating at larger scale. The old adage “don’t do premature optimization” is true here as well. 
  • Norvig-02In general, Google prefers algorithms that get better with data. Not all algorithms are like that, but Google likes to go after the ones with this type of performance characteristic. 

There is no serendipity in Google Search or Google Translate

  • There is no serendipity in search – it is just rehashing. It is good for finding things, but not at all useful for understanding, interpolation & ultimately inference. I think Intelligent Search is an oxymoron ;o)
  • Same with Google Translate. Google Translate takes all it’s cue from the web – it wouldn’t help us communicate with either the non-human inhabitants of this planet or any life form from other planets/milky ways.
    • In that sense, I am a little disappointed with Google’s Translation Engines.  OTOH, I have only a minuscule view of the work at Google.

The future of human-machine & Augmented Cognition

And, don’t belong to the B-Ark !

One Band to Rule them all – Nike FuelBand+ vs. Fitbit Flex


  • I had the Fitbit Flex for a few months. But am not that satisfied with it. Is Nike FuelBand+ any better ? Plan to find out, by wearing them both for a week or two. Am tracking a few factors, pl suggest more …



  • The Good
    • The style & form factor of Nike FuleBand +
      • I am sorry guys, Fitbit is ugly ;o(
    • Fitbit sends you e-mail when the battery is low ! Good work guys
    • Hours won feature of Nike+
      • Good concept. Basically, it wants to you to move periodically
    • Fitbit is a lot lighter than Nike+
  • The Bad
    • Nike Fuel, while a good concept, is a black box. Couldn’t internalize it
    • Hours won notification of Nike+ is useless. It shows on the band’s display, but I miss it most of the time. Others notice the scrolling display !
  • The Ugly
    • The metrics like steps et al from both bands are so different that they can’t be compared
    • Fitbit is maturing. They add more devices in a short span and there is a fundamental difference between the versions. Would have been good if they have an upgrade plan
    • Tracking sleep on both are not that useful

Band Log:

  • Stardate  91454.39 Day 1 : Both devices fully charged
  • Stardate 91461.40 Day 3 : Have been wearing both devices for 3 days. Updating the comparison.
  • Stardate : Updated Summary above.

Comparison: (I will start filling-in in as I go along):

  • Aesthetics – Fashion vs. Functionality:

    • Of course, Fuelband + is a lot good looking.
    • Fitbit should definitely change it’s looks. 
  • Dashboard:

    • First cut, Fitbit looks better, may be because I am use to it.
    • Day 3: Am used to Nike+ dashboard as well. The Nike+ dashboard on iPhone 5S looks very functional
  • Track:

    • Nike has only one main trackable feature NikeFuel against a goal.
    • Fitbit has multiple features – each one with it’s own goals
    • Two Screens after Day 1 below
    • Day-01-fb Day-01-n
    • Nike
      • Nike Fuel Graph : My Nike Fuel goal is 2500. Haven’t yet figured out how it is calculated
      • Hours Won is interesting. Keeps you moving
      • Steps, Calories – Informational. But they are different for the devices.
    • Fitbit
      • Very Active Minutes – 30 min is my goal. Good metric to track
      • Distance, Steps, Calories – Normal metrics
  • Gamification

    • Both devices have badges, trophies, buddy system (friends, groups). I haven’t explored them yet.
    • May be I will buddy with a turtle and feel good ;o)

  • Inactivity

    • Fitbit does nothing except log it
    • Nike has an alert mechanism. Let us see if it works
  • Accuracy:

    • I will jot down counters from both devices. Let us see how they stack up
    • Here are the screen shots for 2 days. I took the shots at the same time.
    • The devices do not agree at all. I think Fitbit is far off on the plus side while Nike might be more closer.
      • This was my first concern with Fitbit and I had contacted their support. Didn’t get a satisfactory response. The support folks are very good, but I think this i a technical fault.
    • Day-02-fb  Day-02-n
    • Day-03-fb  Day-03-n
  • Sleep

    • Fitbit’s sleep tracking is a little awkward. You need to log time went to bed & time woke up. As far as I can tell, if you miss one day it is gone
    • Nike has the session feature, I haven’t yet tested it
  • Battery Life and ease of charging:

    • Fitbit sends an e-mail when the battery is low. Very cute and useful.
    • Let us see what Nike Fuelband+ has in store for us
  • Conclusions after 3 days

    • I think I will go with Nike+
      • Nike is more mature in many ways
      • Nike Fuel is a better motivation than FitBit
      • Overall Nike has a better form factor & a better app
      • Fitbit has better counters & goals for each metric. But they are not cohesive
      • Tracking sleep, while kludgy, is better with Fitbit ! (I never thought I will say this ;o))
      • Fitbit has the premium subscription ($50/yr) that gives more analytics.
        • But am not sure it is worth the price. I think it is an overkill.
        • And Nike has the analytics feature in the base product. Of course, Nike might add a paid feature set
      • Fitbit Flex lacks the display
        • I bought the Fitbit Flex and they have the Fitbit Force. Came out within 3-4 months after they introduced the Flex. I think they should have provided an upgrade path
        • I think Fitbit need at least couple more product revs to add a better display
  • Not so fast !

    • 11/11/13 : Nike+ FuelBand crashed ! I get an page full of error when I connect it – even the Nike web site is crashing ! Looks like the Fuelband crashes the Nikeplus site !
    • NikeError
    • I couldn’t fond a way to report this to Nike. Finally send them a mail via their site & twitter
    • I also contacted the Fitbit guys to see if I can swap out the flex for their force. They released Flex too soon.
    • Let us see how the response is from both the companies …
    • 11/11/13 : Night
      • Heard from both. So support is good from Nike and Fitbit.
      • Nike’s twitter support came back. Mail via web site is stuck somewhere
      • Fitbit (via e-mail) politely refused to swap my Flex for Force.
        • I understand – I bought the device 5 months ago.
      • Nike Plus web site is up; my FuelBand Plus doesn’t crash the site anymore
      • Fuelband SE got reset & reinitialized. It is up & running (Unfortunately we can’t say the same about me ;o( I am static & typing)
      • One good side effect – In the process I discovered where Nike has the manual for FuelBand SE. The links are not that obvious. 

America’s Cup Aftermath: 5 Management Lessons

Americas Cup 01Team USA’s win was very impressive – coming back from an 8-1 deficit, with 2 penalty points, and winning with 8-9 was spectacular, indeed!

Naturally corporate leaders everywhere are asking the questions – How did the team achieve this against an equally versatile and talented opponent? What heroic efforts did the team take ? How did the leads motivate the team to stage such a win ?

Interestingly, simplicity is the key

A thought-provoking article by @juliasulek “America’s Cup: Oracle’s ‘secret weapon’ to pull off historic comeback was largely bluster” has some illuminating answers.

5. Reengineer & Redesign your equipment, processes & services based on real-life learning’s  from the marketplace

During Race 5, Oracle had started strong and was leading but fell apart as it rounded the leeward mark and tanked on the upwind leg

  • They made a few design changes to the 130 foot tall wing sail, adding more curvature down low and more twist on the top.
  • The fine tuning of the win trim control was done over time, incorporating the learnings from the races

4. Reduce to the bare essentials (ie keep your drag minimum)

  • The team realized that they need to reduce the drag. They shorted the central spine of the boat, removing the bowsprit to reduce drag

3. Regroup & Rethink (ie Sail the boat differently)

  • The biggest change, came from the crew

We started sailing the boat differently

Even we were surprised how the subtle changes were so significant in the performance of the boat. But we changed the way the guys actually sailed the boat. It took us three or four days to do that properly

  • Even after making the structural changes after Race 5, the team lost 2 more races. … but they were a little more confident with our speed
  • They achieved a knot more speed after the changes. But the most improvement was the learning by the team that created the winning spirit

2. Relearn

We were able to keep learning. In the end, that was the deciding factor

1. Winning Spirit, Confidence & Pep Talk

  • The Skipper (Jimmy Spithill) maintained confidence that a come back is possible
  • Kinley Fowler, one of their sail trimmers who had injured his back and couldn’t race, entertained the crew each morning with his own Australian brand of pep talk.

Lots of laughs at difficult moments !

And that worked ! Lessons, we all can incorporate into our corporate projects !

[Update 10/6/13] Comments from Justin Tsao : Lesson 0 -To do the Impossible, you must bend the Impossible (rule wise). 

Building London Olympics in A Complex World – Lord Sebastian Coe at TCS Summit Europe

No Powerpoint, no slides – just an insightful, motivating, thought-provoking, engaging conversation …


  • Lord Sebastian Coe, the Olympian, delivered an inspiring speech about the agonies & ecstasies of bringing Olympics to London.
  • If you get a chance to hear from @SebCoe, don’t miss it.
  • The talk was full of insightful observations. I could only capture a few highlights. I will post a link to the video if it becomes available.

The two important questions : How & Why

  • The Why of London Olympics : “Use the London Olympic Games to produce lasting change – especially for young people”
  • The How took them more than 13 years and finally it was a success

Understand whom you are delivering to; and the landscape

  • The stakeholders of London Olympics were the 60 Million people on UK; not the Olympic Committee

Demanding stakeholders are good – the more demanding they are, the more committed they will be to your cause. But you should be constantly communicating with them

You can overcome failures so long as you are brave enough to analyze and understand the contributing factors; you need to be brutally honest. Once the factors are understood, the rest gets a lot easier

Understand what “Great” looks like. The quality of your deliverable depends on the amount of work you put in to crisply define “Great” on whatever you do.

Trivia : Seb Coe has good roots in India. His grandfather is from Delhi !

  • Good IMG_1543
  • Guess what was waiting at my hotel room at the Majestic, Seb Coe’s book !
  • In time for my reading during the long flight back …. I will write a review …
  • And, views from my room at the Hotel Majestic and Cannes …


The Struggle is where greatness comes from

Today, my son (who is just 14) pointed me to an insightful blog by Ben Horowitz of a16z.com.


Couple of interesting observations (Of course, I urge you to read the full blog):

  • Technology business is extremely complex – a game of chess rather than checkers
    • Interestingly another partner at a16z.com, Scott Weiss also referred to a successful entrepreneur “laying out a series of chess moves that reveal an even bigger ambition” in a recent Wall Street blog
  • When you are in a turn, focus on the road not on the walls !

Be strong & you will get through, with a little help + a dash of luck

Finally, I am glad to see that folks like Ben are interesting to the young generation … Kaushik (the son) has been encouraging me to start a company … May be I will …

Ref: Image from Microsoft Gallery

Elon Musk @D11

Elon Musk has an interesting set of view points at D11. The Top 10 highlights:


10. The reason for Tesla was to create a compelling long-range electric car that people would buy.

9. Norway is our largest market. The single biggest purchaser of Tesla cars is an ophthalmologist who lives above the Arctic Circle – and takes an electronic pill ;o)

8.  (About Internet Entrepreneurs) “I recommend that people consider arenas outside of the Internet because there’s a lot of industries (particularly oligopolistic industries need new entrants for innovation) that could use that entrepreneurial talent and the skills that people have learned in creating those companies.”  This inspired Loic Le Muer to say “Let us follow Elon”. I agree ! Also an inspiring read – The Dreamers Dilemma

7.  (Mainstream Tesla-”At profit, in market conditions”) ~3-5 years away, would be about 20% smaller than Model S. 3 Steps – High Price-Low Volume, Mid Price-Mid Volume and Low Price-High Volume. They are in Step 2. 3rd generation is 3-5 years away

6. (Politics) I think we shouldn’t give into the cynicism of politics, we should fight the cynicism. And if we don’t, we’ll get the political system we deserve.

5. (Mark Zuckerberg’s FWD.us)  I supported it initially but I think the methods that were employed was a little too much of the Kissinger-esque realpolitik.

4. (About Hyperloop) There’s a Tesla announcement around June 20 … at some point after that will be a good time to talk about it. For those that aren’t aware, the basic idea is will there be a better way to travel quickly from LA to San Francisco than high-speed rail - a “cross between a Concorde, a Railgun & an Air Hockey Table”

3. (About “fully & rapidly” reusable rockets) The cost of the propellent in the rocket is only .3 percent of the vehicle (~$200,000 vs $60,000,000). Think about a commercial aircraft — you don’t want to buy a new plane every time you book a flight.

2. (Spacex) SpaceX was able to achieve orders of magnitude savings in rockets. Instead of looking at what other rockets cost, they looked at the material cost (which is only 1-2% of total cost) which was small; clearly people were doing silly things on how the rockets were put together; for SpaceX, the savings came from efficiently putting the rocket together! Interesting. Elan expected SpaceX to fail !

1.a. Either we spread Earth to other planets, or we risk going extinct. An extinction event is inevitable and we’re increasingly doing ourselves in.

1. “Mars is a fixer-upper of a planet, but we could make it work” – multi planetary life with a base camp in Mars (“difficult but achievable”) ! That is worth living for (“The future is better than the present”)… even though the trip will take 3-6 months … one way … but the security lines should be shorter …

Now is a good time to read the Live Blog and/or watch the Video !

[Update June 3,2013] A summary of D11 observations – good read


Photo: Asa Mathat/AllThingsD

TCS Siruseri Campus

Today am visiting the TCS Siruseri campus in Chennai. Very elegant & interesting structure built by an Urguan Architect.

  • The legend has it that while discussing the architecture, Carlos drew butterflies & that stuck as the theme.
  • A majestic 5-floor-high open Atrium corridor forms the spine of the butterfly with 6 buildings forming the wings, the buildings themselves are butterflies w/ a small spine elevator bank in the middle and the two north & south wings buildings.
  • The side view below shows three buildings EC1(right),EC2 & EC3. I am in EC3.
  • The spine atrium on the far side has a pond, benches, shops & interesting restaurants – The Saravana Bhavan food is exotic. The spine even has a Subway !
  • The campus hosts > 20,000 associates

The second picture below shows the full 6 buildings and the observation tower (which is still under construction)

image description



[Update 1 June 3, 2013] Collection of pictures that take you on a tour to other TCS Campuses.

Kareem Abdul-Jabbar: 20 things I wish I had known when I was 30

Kareem-Abdul-Jabbar-Skyhook-WallpaperKareem Abdul-Jabbar has an excellent blog at Esquire on 20 pieces of advise to the younger self at 30. Thanks to Jason Hiner‘s tweet.

The blog & the comments are a must read.

A few of them hit home for me:

  • Be patient
  • Listen More than Talk
  • Being right is not always the right thing to be
  • Do one thing every day that helps someone else.
  • Do one thing every day that you look forward to doing. 
  • Don’t be so quick to judge.
  • Everything doesn’t have to be fixed.
  • Play the Piano
  • Become Financially Literate

Ref: Wallpaper from http://www.basketwallpapers.com/USA/Kareem-Abdul-Jabbar/

Is our Neocortex a Giant Semantic Bloom Filter ? Of Natural Intelligence, Machine Learning & Jeff Hawkins



In a set of four lectures spanning about 3 years, Jeff Hawkins explains how & why big data can only be solved by evolutionary-adaptive-continuously-learning models incorporating principles from the working of Neocortex.
It does make sense – especially for NLP, NLU & Knowledge Representation. I am a big fan of the Borgs and their coordinated intelligence.

These are my annotated picture-notes …


Let me begin at the beginning. The other day I came across 4 very interesting talks by Jeff Hawkins on Biological Inspired Machine intelligence.

Call it serendipity because we have been looking for more effective ways for Knowledge Representation (KR) & Natural Language Understanding (NLU)

For example movie names, while very easy for humans to understand, a MaxEnt NER finds it very hard.  Knowledge Representation & Association is more harder !

We are experimenting with a few techniques like word-based tries (ie. spell-check sentences by words), higher order federated Bloom Filters and n-gram hashing. Planning to incorporate some of Jeff’s ideas …

I digress … Topics for another day … back to Jeff & Machine Intelligence …

Very inspiring, extremely thought provoking talks – as usual the inimitable Jeff Hawkins at his best

  1. Google Tech Talk : Jeff Hawkins, “Building Brains to Understand the World’s Data
  2. UC Berkeley Graduate Lectures
  3. “Advances in Modeling Neocortex and its impact on Machine Intelligence” by Jeff Hawkins,  Smith Group Lecture presented at the Beckman Institute for Advanced Science & Technology at the University of Illinois at Urbana-Champaign

Le Plat Principal:

The four talks have lot of depth and are packed. Moreover Jeff talks very fast – I listened to the talks a few times – at least 3 hrs per one hour talk. You should listen to them slowly & rewind as reqd. It takes a few hours to get one’s head around the various ideas.

Let me annotate a few of his slides – those I was able to internalize to some extent:

Focus & premise[3]:


The assertion, that many problems can only be solved by incorporating principles from the working on Neocortex, is interesting.

BTW, it does make sense – especially for NLU & Knowledge Representation.

As Jeff mentions later, the behavior need not be human-like, but the representation, interpretation & “understanding” would be.

Neocortex Architecture[3]:

“Neocortex is just a sheet of cells  2mm thick, the size of a dinner napkin” – Amazing what it can do!


The Six Principal Essentials of Biological Intelligence

The picture says it all.


Learning involves training and adaptive connections


The concept of streaming events & the learning mechanisms

Patterns from complex data streams


The paper “Hierarchical Temporal memory” has the gory details about the Hierarchical Temporal Learning.



Interesting observation: Emotion, the fundamental aspect of being human, is not a requirement for intelligence – reminds us of Spock, of course.

Machine intelligence is not about replicating human behavior or even passing the turing test. I agree on this – we need the machines to think & do things we cannot do thus augmenting us. Make us stronger where we are weak !

Le Digestif

What interested me most was the sematic knowledge representation, NLP & NLU. The ability to understand and store concepts, the capacity to generalize as well as the mechanisms of strengthening and weakening connections based on external signals – just beautiful …

Agree that the Sparse Distributed Representation could be the language of all the intelligent machines.

The SDR looks a lot like a giant Bloom Filter


Hawkins-100-11-01The planes can be considered as rows and a column as the temporal dimension of the semantic mapping (the memory of sequences). Which equates to a giant n-dimensional Bloom Filer – a data structure we can grok (Pun intended as Jeff’s product is called Grok!).

The bloom filter analogy, while extremely simplistic, is conceptually congruent, in the sense that “similar values have similar representation”, of course depending on the hash algorithm.

After listening to the talks and thinking them over, I have a thousand questions in many directions. I will post the answers as we develop this through for our needs. Please send in your insights as comments to this blog. AM sure it will help a few folks !


  1. How do we handle semantic categories ? 
  2. How do we build more sophisticated representations based on spatial patterns ?
  3. What is the hash function that maps a slice of semantic to this giant Bloom Filter ?
  4. How does it handle collision? Corruption ? Clustering for resiliency/self adjusting representation ?
    • Collision might be good and I think that is what Jeff calls as semantic generalization
  5. How does the semantic slice mapping function differentiate between a search & computation to trigger appropriate actions?
    • For example the following two questions require different actions: 
      • What is stock price of IBM ?” vs.
      • What is the volatility as reflected in the beta of IBM for this quarter ?” 
      • The first one is a search while the second has computation …
  6. Is the hash function same for all of us or is it different for each person ?
    • Most probably the function is a learned artifact.
  7. Another interesting vector is the Hierarchy & higher patterns of temporal coalescence/slowness – the high-order capability, tweaking the learning rates across the layers.
    • How can this be modeled with the analytical data structures we have?
    • And what are the mechanics for stable representation of pattern sequences – because with dynamicity and temporality comes the difficulty of snapshots and consistency between them.
    • The unique representation of the same sequence, at a later time in context of the earlier invocation is interesting …
  8. How do we “put a classifier on the top” ?
    • Play with permanence? Probability?
  9. What are the algorithms to prevent run away prediction?
    • I agree that we could account for rapid state difference vs. slower state; we still will have to encapsulate it in some form of code

Finally, can we build “Amazingly Intelligent Machines?” Yes We can !

And agree with Jeff that “It is essential, for the survival of the spices, that we build them” …