Posted on

A great startup ideas

  • they’re something the founders themselves want.
  • that they themselves can build.
  • that few others realize are worth doing.
  • Its unsexy
  • There is a lot of shit you will need to shovel to get things to work
  • It requires persistent and dedication beyond a time horizon that most folks are willing to devote their time to
  • A small number of folks chronically in painful need of the solution
  • There are other folks out there with shitty solutions in the market

External references

Posted on

Signals on Leadership

  • 10 essentials of Peter Drucker – liberal arts and leadership
  • Jeff Bezos, Amazon –  13 principles of leadership
  • Genghis Khan – clearly defined rules
Posted on

Book summary – The inmates are running the asylum

  • Cheaper in the long run to be slow but build the proper user experience
  • Fixating in a schedule and being forced to take short cuts without a clear vision in mind is bad
  • Cheaper to figure out the experience before building it. The latter is like scalping with a chainsaw
  • Code revisions due initial lack of clarity in the market inevitably scars the code
  • listen to your users but use your own judgement to discern the product feature to build
  • handing the reins over to the users results in the abdication of your vision
  • abdication of your vision pulls you to the mainstream and results in the loss of differentiation
  • if they believe if you vision they will be willing to wait for it anyway
  • reality is strong enough to stand on its own feet. It does not need a devils advocate
  • Distinguish clearly between task and goals
  • goals are universal across time while tasks are transient
    • goals – company needs to make money
    • tasks – need to write a report in windows power point
  • there are personal goals and professional goals
    • personal goal – feel good and don’t look stupid
    • professional goals – make sure boss gets his report and I don’t get fired
Posted on

An afternoon with writers Terry Books and Tad Williams

   – fiction is about leaving enough space for the user to apply her imagination
   – reading is a very engaging activity between the writer and the reader
   – fiction allows you to write about the real world without having to exactly mirroring it
   – America is an extremely polarized country. Same theme elicits different responses
   – the writer needs to be able to step out of his own perception into another perception that might be opposed
    – once books gets turned into show sales of books go way through the roof
   – to only focus attention on things you have control of
    – works building happens in the background through a process of collecting random thoughts – JRR Tolkien spent 40-50 years working on the lord of the rings
    – process changes as time in life stage changes
    – daily circadian rhythm dictates the writing schedule mornings used for writing
    – science and magic are symbolism of power
    – more stimulated by reading outside the field
        – start by reading within the field and understand what works
        – then sent over by the editor or if knows the author
        – wants to read book that impacts us in a meaningful way
    – power results in a lot of decision that needs to be made night and day. Causes a lot of aging really fast
    – doing something for fun feels quite different from working full time
    – writing is like raising a child. You raise them and then send them out into the world to make money
    – if I can’t write a scene I need to
    – if I can find an answer within 10 questions then it’s probably not worth writing
   – need to think about an idea till invested before committing to it
    – it’s like a survival of the fittest with ideas in the head
    – make sure to marry and partner with the right person. It’s unbelievable what they have to deal with
Posted on

The surest cure to all forms of narrative fallacy

The problem with the limbic brain is that it is too susceptible to narratives. To the extent it gets bogged down by narratives of any sorts, the person losses its ability to fluidly respond to arising circumstances.

The only way out is to mitigate the binding effects of narratives. The surest way to do so is by short circuiting all lesser narratives by the most fundamental of narratives and realigning all resource allocation backwards from that single North Star. Your imminent Death. Yes, in case you don’t remember your biological body will inevitably become depleted of life force after a fixed number of biochemical cycles.

Death is your best friend, it forces you to figure out your deepest most fundamental truth. In the face of Death, all other lesser narratives fade away. In the face of Death, there is no longer any inhabitation. In the face of Death, you gain the ultimate Freedom to expend your remaining biochemical cycles based on your core fundamental truth.

Death is your best friend who watches your back day and night. It is also the best antidote to all marketing, propaganda and bullshit.

Posted on

And the trends does a sharp turn

Looking at the reports today, I realized to myself how easy it is for us to succumb to our brain’s inherent biases – in this case the narrative fallacy / apophera

Just last week, I noticed my anxiety slowly creeping up as for days in row, there were a total lack of buying opportunities that got surfaced by the report because every stock was going up.

Day after day, as the trend continued self doubts started creeping in slowly. Stories of myself sitting out because of some key factors I missed in my model while the largest bull run ever happened constantly haunted my thoughts.

And then I did the unthinkable on Monday and succumb to this narrative that was going on in my head. I deviated from my strategy only to pick a 10% capital loss within a day before coming to my senses.

Retaining the ability to sit and do nothing while the whole world gets taken over by hype or acting when everyone is cowering in fear is the hardest form of discipline to practice.

 

 

Posted on

Key takeaways from presentation by Hannes

Cosine similarity
Simhash distance
https://github.com/facebookresearch/StarSpace

Most computers can do 0.5 Terraflops per second
needs to wait 30 seconds to do recommendation

GPU can do multiple terraflops per second
Need to figure out cheaper way to do so

Neural network computing should be on the phone in the future
Mobile net

Posted on

The Web is a giant graph database.

I was trying for the longest time to articulate the problem I was trying to solve when the most suitable analogy finally occurred to me when I was doing my evening meditation just this evening.

The web is a giant graph database where each web page is a node in the graph.

Just like any node within a graph database, it contains attributes. In this case, there are two types of attributes:

  1. values embedded within the DOM elements of the page
  2. values embedded within the URL of the page

Just like any relationships between nodes within a graph database, each node can be related to other nodes. In this case, the relationships can be expressed in two forms:

  1. explicit hyperlinks from other page to another
  2. attribute of a node who value matches the attribute that exist in another node

People intuitively know the existence of this giant graph database, but few actually see it clearly.

Some folks who “own” a huge number of these nodes try to prevent or make it hard for others to access the attributes embedded within the nodes they own. It is a good thing that governments around the world are starting to realize that some of these nodes don’t really belong to these folks but really belong to people of which the attributes within these nodes describe.

Other folks seeing the value of some of the attributes within some nodes try to gather these attributes through this activity called web scraping. The problem with their usual approach is there they don’t try to solve the problem at a fundamental level but instead resort to a cheap one-off kind of approach. The resultant solution is usually a tight coupling between schematic declaration and code. Imagine mixing HTML code with server side code. It becomes spaghetti code really fast. It is really hard to maintain and does not promote reusability.

From my point of view, it is not really their fault these folks generate spaghetti code. The heart of the problem is the missing Semantic Query Language that in normal circumstances should have come accompanied with the database which would have allowed developers the ability to query for data from this giant graph database without the need to concern himself with the underlying implementation details.

And that is the problem I am trying to solve. Defining the missing Semantic Query Language as well as building the corresponding ODBC engine that would interpret the query defined by the user.

The mission for GetData.IO thus is to build an ODBC layer so as to allow anybody to query for data from this giant graph database called the Web. We need to ensure they can easily do so be it via a simple point and click interface or just writing the Semantic Query by hand.

They should not be forced to pay horrendous amount of money to web-scraping companies to run this query on their behalf. They should not be forced to consume in what ever form intermediaries see fit to present the data in. They should have the freedom to query data from this giant graph database in whatever way they deem fit.

We want to democratize access to this giant graph database of human knowledge.

Posted on

Reflections for the evening

In a sufficiently liquid market where talent is in short supply and jobs are abundant, there are more upsides then downsides in swinging for the fences by undertaking moonshot projects. The fact that it is such a crazy idea inevitably causes it to be awe-inspiring as a by-product. It also happens to make it easier to raise money and attract talents.

The main pre-requisite for risk taking is the ability to stomach uncertainty and ambiguity. Lessons learned from Michael Porter on strategy, Reed Hastings and my favorite childhood game “Uncharterd Waters” by Koei, dictates that success is best achieved when your entire operations’ configuration is strictly geared towards your end goal so as to eliminate wastage sin the form of misaligned resource deployment.

In a sufficiently dense professional network, being a highly trusted member of good standing known for delivering high quality work inevitably draws opportunity and resources. This support facilitates higher levels of risk taking than if operating in a territory where the network is non-existent.

During Seth Godin’s Interview, he mentioned a few very important points:

  • It is important to get comfortable with feeling incompetent because considering the ever shifting demand, that is where the largest area of growth is available
  • most people join big company because they want to feel confident by being able to focus only on things they are good at.
  • Trust is the most important currency you can invest in
  • The best way to crack product market fit is by winning the trust of a very niche and well-defined set of people. It should be small enough that you almost feel insecure about the market size. 10 is a very good number.
  • You want to make them so happy they want to tell other people about your product/service. People who have similar needs generally know each other.
  • Humilty and Empathy are the most important character traits to cultivate if you want to win trust.
  • prioritize resources towards building trust as opposed to advertising in the initial phase of your project