Tag: analysis

Never Going Back

Third-Order Consequences
This is where things start getting really wild. Unshackled by the office’s location, people will move where they want to move, often out of expensive cities and into more affordable towns with better weather. Those are first-and second-order effects. Some third-order effect might be that because they’re saving more money, they have more money to invest, and the trends that I wrote about in Software is Eating the Markets towards more retail investment in stocks, art, real estate, and more will accelerate.

Other potential third-order effects include:

Rise of Alternative Education. As mobility increases, more people will need to give online or alternative education a real shot, because they’ll be loath to gain freedom from the office but remain tied down by their childrens’ schools. Homeschooling options like SchoolHouse, which matches groups of families with teachers to form microschools, Primer, an online homeschooling community and infrastructure startup, or Outschool, which lets kids take online classes or camps from anywhere, will appeal to parents who want to move while keeping their kids well-educated.

More Fluid Employment. Productive employees may work multiple full-time roles. In GitLab’s fourth phase, Intentionality, employees are measured on output. If employees can keep up the output, employers will be comfortable letting them work multiple jobs. The absolute star performers, who Dror calls “The 10x Class,” will put their talents up for a global auction, and will reach income levels similar to top athletes and celebrities.

New Employee Stock Options. As companies and employees enjoy a more transitory relationship, and as Remote leads to more precise performance tracking, equity will have to evolve to be rewarded for performance and contribution instead of tenure and rank. As Sari and I wrote, we think that Fairmint is in a great position to make this possible technically. Remote will make it acceptable culturally.

Organizational Metaphors

NOBL Academy
Machine: an organization is a series of connected parts arranged in a logical order in order to produce a repeatable output
Organism: an organization is a collective response to its environment and, to survive, must adapt as the environment changes
Brain: an organization is a set of functions designed to process information and learn over time
Cultural System: an organization is a mini-society, with its own culture and subcultures defined by their values, norms, beliefs, and rituals
Political System: an organization is a game of gaining, influencing, and coordinating power
Psychic Prison: an organization is a collection of myths and stories that restrict people’s thoughts, ideas, and actions
Instrument of Domination: an organization is a means to impose one’s will on others and exploit resources for personal gains
Flux and Transformation: an organization is an ever-changing system indivisible from its environment

100 Little Ideas

A list of ideas, in no particular order and from different fields, that help explain how the world works.

  1. Depressive Realism
  2. Skill Compensation
  3. Curse of Knowledge
  4. Base Rates
  5. Base-Rate Neglect
  6. Compassion Fade
  7. System Justification Theory
  8. 3 Men Make a Tiger
  9. Burdian’s Ass
  10. Pareto Principle
  11. Sturgeon’s Law
  12. The Matthew Effect
  13. Impostor Syndrome
  14. Anscombe’s Quartet
  15. Ringelmann Effect
  16. Semmelweis Reflex
  17. False-Consensus Effect
  18. Boomerang Effect
  19. Chronological Snobbery
  20. Outgroup Homogeneity
  21. Planck’s Principle
  22. McNamara Fallacy
  23. Courtesy Bias
  24. Berkson’s Paradox
  25. Group Attribution Error
  26. Baader-Meinhof Phenomenon
  27. Ludic Fallacy
  28. Normalcy Bias
  29. Actor-Observer Asymmetry
  30. The 90-9-1 Rule
  31. Texas Sharpshooter Fallacy
  32. Fredkin’s Paradox
  33. Poisoning the Well
  34. Golem Effect
  35. Appeal to Consequences
  36. Plain Folks Fallacy
  37. Behavioral Inevitability
  38. Apophenia
  39. Self-Handicapping
  40. Hanlon’s Razor
  41. False Uniqueness Effect
  42. Hard-Easy Effect
  43. Neglect of Probability
  44. Cobra Effect
  45. Braess’s Paradox
  46. Non-Ergodic
  47. Pollyanna Principle
  48. Declinism
  49. Empathy Gap
  50. Abilene Paradox
  51. Collective Narcissism
  52. Moral Luck
  53. Feedback Loops
  54. Hawthorne Effect
  55. Perfect Solution Fallacy
  56. Weasel Words
  57. Hormesis
  58. Backfiring Effect
  59. Reflexivity
  60. Second Half of the Chessboard
  61. Peter Principle
  62. Friendship Paradox
  63. Hedonic Treadmill
  64. Positive Illusions
  65. Ironic Process Theory
  66. Clustering Illusions
  67. Foundational Species
  68. Bizarreness Effect
  69. Nonlinearity
  70. Moderating Relationship
  71. Denomination Effect
  72. Woozle Effect
  73. Google Scholar Effect
  74. Inversion
  75. Gambler’s Ruin
  76. Principle of Least Effort
  77. Dunning-Kruger Effect
  78. Knightian Uncertainty
  79. Aumann’s Agreement Theorem
  80. Focusing Effect
  81. The Middle Ground Fallacy
  82. Rebound Effect
  83. Ostrich Effect
  84. Founder’s Syndrome
  85. In-Group Favoritism
  86. Bounded Rationality
  87. Luxury Paradox
  88. Meat Paradox
  89. Fluency Heuristic
  90. Historical Wisdom
  91. Fact-Check Scarcity Principle
  92. Emotional Contagion
  93. Tribal Affiliation
  94. Emotional Competence

TikTok and the Sorting Hat

TikTok doesn’t bump into the negative network effects of using a social graphs at scale because it doesn’t really have one. It is more of a pure interest graph, one derived from its short video content, and the beauty is its algorithm is so efficient that it its interest graph can be assembled without imposing much of a burden on the user at all. It is passive personalization, learning through consumption. Because the videos are so short, the volume of training data a user provides per unit of time is high. Because the videos are entertaining, this training process feels effortless, even enjoyable, for the user.

the reason tiktok has taken off is because it’s recommendation algorithm is really good, and it doesn’t need a social graph to thrive.

Solving online events

It’s often struck me that networking events are pretty inefficient and random. If you’re going to spend 1 hour or 2 in a room with 50 or 500 people, then you could take that as a purely social occasion and enjoy yourself. But if your purpose is to have professionally useful conversations, then what proportion of the people in the room can you talk to in 1 hour and how likely is it that they’ll be the right ones? Who’s there? I sometimes suggest it would be helpful if we all wore banners, as in the image at the top, so that you could look across the room and see who to talk to. (First Tuesday did something like this in 1999, with different colored badges.)

This might just be that I’m an introvert asking for a machine to manage human connections for me (and I am), but there is also clearly an opportunity to scale the networking that happens around events in ways that don’t rely on random chance and alcohol tolerance. A long time ago Twitter took some of that role, and the explosion of online dating also shows how changing the way you think about pools and sample sets changes outcomes. In 2017, 40% of new relationships in the USA started online. Next, before lockdown, you would often have planned to schedule a non-urgent meeting with a partner or client or connection ‘when we’re in the same city’. That might be at some specific event, but it might also just be for some ad hoc trip – ‘next time I’m in the Bay Area’ or ‘next time you’re in New York’. In January most people would never actually have thought of making that a video call, but today every meeting is a video call, so all of those meetings can be a video call too, and can happen this week rather than ‘next time I fly to that city’ – or ‘at CES/NAB/MIPCOM’. In the last few months video calls have broken through that habit. I wonder what happens if we accelerate all of those meetings in that way.

On the unbundling of events, and how networking might be done better.

On GPT-3

GPT-3 is scary because it’s a tiny model compared to what’s possible, with a simple uniform architecture trained in the dumbest way possible (prediction of next text token) on a single impoverished modality (random Internet text dumps) on tiny data (fits on a laptop), and yet, the first version already manifests crazy runtime meta-learning—and the scaling curves still are not bending! The samples are also better than ever, whether it’s GPT-3 inventing new dick jokes or writing (mostly working) JavaScript tutorials about rotating arrays. Does it set SOTA on every task? No, of course not. But the question is not whether we can lawyerly find any way in which it might not work, but whether there is any way which it might work.