{
  "version": "https://jsonfeed.org/version/1",
  "title": "Madhav Singhal",
  "home_page_url": "",
  "feed_url": "https://msinghal.dev/feed.json",
  "description": "Madhav Singhal's Personal Website",
  "icon": "https://msinghal.dev/assets/apple-touch-icon.png",
  "favicon": "https://msinghal.dev/assets/favicon.ico",
  "expired": false,
  "author": {
    "name": "Madhav Singhal",
    "url": "https://msinghal.dev/"
  },
  "items": [
    
    

    
    {
      "id": "162683ba931681b1b31f7adadff0c3b67d29b832",
      "title": "What I Have Learnt So Far",
      "summary": "What I have learnt, both intellectually through consuming other people’s ideas and viscerally through my own experiences.",
      "content_text": "Each of the following is something non-obvious that I have learnt both \u0026lsquo;intellectually\u0026rsquo;, i.e. through consuming other people’s ideas, and \u0026lsquo;viscerally\u0026rsquo;, i.e. through my own experiences. I believe that these ideas are worth sharing:\n What you work on is more important than what you know, how hard you work and how capable you are. In choosing work, work only at the frontier where things are being rapidly discovered, invented or created. Everything else is a suboptimal use of time and effort for creating and capturing intellectual and/or economic value. Know what you want with high clarity and optimize all decisions around it. If you are not getting something you want, you are getting the input variables wrong. The context is more important than the agent. Change your environment before changing yourself. That which is \u0026lsquo;n of 1\u0026rsquo; is far more valuable than that which is a winner in a category with ‘n \u0026gt; 1’. Things often succeed because they are sufficiently unique rather than sufficiently good. To sell yourself, all sales have to be inbound. Good communication is explicating the structure of the message in the message itself. Good writing often resembles someone talking, allowing the reader to easily hear it as a voice in their head. The hypothesis matters more than experiment and data. The most important factor in choosing family, friends, and co-workers is whether your core values align. Know the distinction between signal and substance, and identify where each one is being used to assess value. Act accordingly. Most mispricing occurs because of a divergence between signal and substance, and most rejections occur because you couldn’t meet a required threshold of either. The singular explainer and influencer of human behavior is incentive. Your value in a group is often determined by your value outside the group. The elephant in the brain is very real: people often unconsciously misperceive reality and the reasons for their actions, creating fabricated justifications for their own benefit. All opportunities have information asymmetries. You need to be on the right side of them. Truth is a local variable. Risk is a necessary condition to create value, and your operating philosophy should be to consistently take non-blow up risk. What you take a risk on matters more than the amount of risk you take. Often, the arrow of causation is in the reverse direction than what it seems, so always think in both directions. “Your life is a series of projects”. Projects are public interfaces to your skills. Focus on accumulating projects more than accumulating skills. Wealth exists across 4 kinds of capital: financial, social, career, and knowledge. You should be compounding each aggressively. The mechanics of compounding are feedback loops and phase shifts. Most outcomes of compounding look like they happen “slowly, then all at once”. To remember what you are learning, incessantly ask “how does this connect to what I already know?” Learn with a barbell: learn the fundamentals and the latest discoveries and inventions; read the textbooks and read the latest papers. For learning anything there is immense value in just “pretraining” — just keep reading and consuming information on it. Starting with your intended application and recursively learning what you need is far more effective than learning first and applying later. Clear understanding and explanations often boil down to being mechanistic — a simple chain of reasoning across objects, their properties and the relationships between them. Framing the problem is the most critical step to solving it. Iterative problem solving is just iterative problem framing. Ideas not acted upon effectively end up being non-existent. The rarest but highest ROI human interaction is encouragement. Encouragement makes people perform way beyond their capacity. It is actually valuable to be mimetic in certain situations and you need to figure out when to be mimetic and when to be not. It’s not what you say, it’s what they hear. You are optimizing for the latter. Your job is not to follow rules and criteria. Your job is to become so good that rules and criteria don’t apply to you. Words of wisdom don’t come from the weak. Who you are determines the value of what you say. If the weekends look different from the weekdays, there is something wrong with “work”. You tend to see more risk in ‘risky’ opportunities and less risk in ‘safe’ opportunities. What you enter a problem with is less important than your velocity of iteration and you figuring it out as you go. The packaging is as much the product as the product itself. Be surgical about your own packaging. Everyone has a brand that they are always optimizing for, and you should figure how you fit in if you want something from them. The halo effect is a real, powerful signal. You need to make it work for you. It’s easier to do harder things in the long run. What seems easy at the get go sometimes has hidden difficulties or outright failure guarantees. You are simply a machine: program yourself overtime with the right inputs, you will get the outputs you want. Doing the thing directly is far more useful than doing things about the thing. Ironically, the most important trait for clear thinking is to hold contradictory ideas in your head simultaneously. The most important but often overlooked thing in business is what the customer wants. Capitalism rewards outcomes not inputs, efforts or outputs. Advice is only useful if there is sufficient applicability of context. Any advice given devoid of context is noise. That which is surprising to you has the highest informational value for you. What you do affects who you are, not who you are affects what you do.  \nThanks to Arth Gupta and Rahul Garg for their feedback on this post.\n",
      "content_html": "\u003cp\u003eEach of the following is something non-obvious that I have learnt both \u0026lsquo;intellectually\u0026rsquo;, i.e. through consuming other people’s ideas, and \u0026lsquo;viscerally\u0026rsquo;, i.e. through my own experiences. I believe that these ideas are worth sharing:\u003c/p\u003e\n\n\u003col\u003e\n\u003cli\u003eWhat you work on is more important than what you know, how hard you work and how capable you are.\u003c/li\u003e\n\u003cli\u003eIn choosing work, work only at the frontier where things are being rapidly discovered, invented or created. Everything else is a suboptimal use of time and effort for creating and capturing intellectual and/or economic value.\u003c/li\u003e\n\u003cli\u003eKnow what you want with high clarity and optimize all decisions around it.\u003c/li\u003e\n\u003cli\u003eIf you are not getting something you want, you are getting the input variables wrong.\u003c/li\u003e\n\u003cli\u003eThe context is more important than the agent. Change your environment before changing yourself.\u003c/li\u003e\n\u003cli\u003eThat which is \u0026lsquo;n of 1\u0026rsquo; is far more valuable than that which is a winner in a category with ‘n \u0026gt; 1’. Things often succeed because they are sufficiently unique rather than sufficiently good.\u003c/li\u003e\n\u003cli\u003eTo sell yourself, all sales have to be inbound.\u003c/li\u003e\n\u003cli\u003eGood communication is explicating the structure of the message in the message itself.\u003c/li\u003e\n\u003cli\u003eGood writing often resembles someone talking, allowing the reader to easily hear it as a voice in their head.\u003c/li\u003e\n\u003cli\u003eThe hypothesis matters more than experiment and data.\u003c/li\u003e\n\u003cli\u003eThe most important factor in choosing family, friends, and co-workers is whether your core values align.\u003c/li\u003e\n\u003cli\u003eKnow the distinction between signal and substance, and identify where each one is being used to assess value. Act accordingly. Most mispricing occurs because of a divergence between signal and substance, and most rejections occur because you couldn’t meet a required threshold of either.\u003c/li\u003e\n\u003cli\u003eThe singular explainer and influencer of human behavior is incentive.\u003c/li\u003e\n\u003cli\u003eYour value in a group is often determined by your value outside the group.\u003c/li\u003e\n\u003cli\u003eThe elephant in the brain is very real: people often unconsciously misperceive reality and the reasons for their actions, creating fabricated justifications for their own benefit.\u003c/li\u003e\n\u003cli\u003eAll opportunities have information asymmetries. You need to be on the right side of them.\u003c/li\u003e\n\u003cli\u003eTruth is a local variable.\u003c/li\u003e\n\u003cli\u003eRisk is a necessary condition to create value, and your operating philosophy should be to consistently take non-blow up risk.\u003c/li\u003e\n\u003cli\u003eWhat you take a risk on matters more than the amount of risk you take.\u003c/li\u003e\n\u003cli\u003eOften, the arrow of causation is in the reverse direction than what it seems, so always think in both directions.\u003c/li\u003e\n\u003cli\u003e“Your life is a series of projects”. Projects are public interfaces to your skills. Focus on accumulating projects more than accumulating skills.\u003c/li\u003e\n\u003cli\u003eWealth exists across 4 kinds of capital: financial, social, career, and knowledge. You should be compounding each aggressively.\u003c/li\u003e\n\u003cli\u003eThe mechanics of compounding are feedback loops and phase shifts. Most outcomes of compounding look like they happen “slowly, then all at once”.\u003c/li\u003e\n\u003cli\u003eTo remember what you are learning, incessantly ask “how does this connect to what I already know?”\u003c/li\u003e\n\u003cli\u003eLearn with a barbell: learn the fundamentals and the latest discoveries and inventions; read the textbooks and read the latest papers.\u003c/li\u003e\n\u003cli\u003eFor learning anything there is immense value in just “pretraining” — just keep reading and consuming information on it.\u003c/li\u003e\n\u003cli\u003eStarting with your intended application and recursively learning what you need is far more effective than learning first and applying later.\u003c/li\u003e\n\u003cli\u003eClear understanding and explanations often boil down to being mechanistic — a simple chain of reasoning across objects, their properties and the relationships between them.\u003c/li\u003e\n\u003cli\u003eFraming the problem is the most critical step to solving it. Iterative problem solving is just iterative problem framing.\u003c/li\u003e\n\u003cli\u003eIdeas not acted upon effectively end up being non-existent.\u003c/li\u003e\n\u003cli\u003eThe rarest but highest ROI human interaction is encouragement. Encouragement makes people perform way beyond their capacity.\u003c/li\u003e\n\u003cli\u003eIt is actually valuable to be mimetic in certain situations and you need to figure out when to be mimetic and when to be not.\u003c/li\u003e\n\u003cli\u003eIt’s not what you say, it’s what they hear. You are optimizing for the latter.\u003c/li\u003e\n\u003cli\u003eYour job is not to follow rules and criteria. Your job is to become so good that rules and criteria don’t apply to you.\u003c/li\u003e\n\u003cli\u003eWords of wisdom don’t come from the weak. Who you are determines the value of what you say.\u003c/li\u003e\n\u003cli\u003eIf the weekends look different from the weekdays, there is something wrong with “work”.\u003c/li\u003e\n\u003cli\u003eYou tend to see more risk in ‘risky’ opportunities and less risk in ‘safe’ opportunities.\u003c/li\u003e\n\u003cli\u003eWhat you enter a problem with is less important than your velocity of iteration and you figuring it out as you go.\u003c/li\u003e\n\u003cli\u003eThe packaging is as much the product as the product itself. Be surgical about your own packaging.\u003c/li\u003e\n\u003cli\u003eEveryone has a brand that they are always optimizing for, and you should figure how you fit in if you want something from them.\u003c/li\u003e\n\u003cli\u003eThe halo effect is a real, powerful signal. You need to make it work for you.\u003c/li\u003e\n\u003cli\u003eIt’s easier to do harder things in the long run. What seems easy at the get go sometimes has hidden difficulties or outright failure guarantees.\u003c/li\u003e\n\u003cli\u003eYou are simply a machine: program yourself overtime with the right inputs, you will get the outputs you want.\u003c/li\u003e\n\u003cli\u003eDoing the thing directly is far more useful than doing things about the thing.\u003c/li\u003e\n\u003cli\u003eIronically, the most important trait for clear thinking is to hold contradictory ideas in your head simultaneously.\u003c/li\u003e\n\u003cli\u003eThe most important but often overlooked thing in business is what the customer wants.\u003c/li\u003e\n\u003cli\u003eCapitalism rewards outcomes not inputs, efforts or outputs.\u003c/li\u003e\n\u003cli\u003eAdvice is only useful if there is sufficient applicability of context. Any advice given devoid of context is noise.\u003c/li\u003e\n\u003cli\u003eThat which is surprising to you has the highest informational value for you.\u003c/li\u003e\n\u003cli\u003eWhat you do affects who you are, not who you are affects what you do.\u003c/li\u003e\n\u003c/ol\u003e\n\n\u003cp\u003e\u003cbr\u003e\n\u003cbr\u003e\n\u003cbr\u003e\n\u003cbr\u003e\n\u003cbr\u003e\u003c/p\u003e\n\n\u003chr /\u003e\n\n\u003cp\u003eThanks to Arth Gupta and Rahul Garg for their feedback on this post.\u003c/p\u003e\n",
      "url": "/2024/01/04/what-i-have-learnt/",
      "image": "photos/<no value>",
      "banner_image": "photos/<no value>",
      "date_published": "4016-04-09T10:44:00+00:00",
      "date_modified": "4016-04-09T10:44:00+00:00",
      "author": {
        "name": "Madhav Singhal",
        "url": "https://msinghal.dev/"
      }
    }
    
  ]
}