notes/Small-groups (Link Bibliography)

“notes/​Small-groups” links:



  3. ⁠, Alex Danco (2019-11-27):

    The difference in angel investing between Silicon Valley and everywhere else isn’t just a difference in perceived risk/​​​​reward or a difference in FOMO. It’s that angel investing fulfils a completely different purpose in Silicon Valley than it does elsewhere. It’s not just a financial activity; it’s a social status exercise.

    Angel Investors in the Bay Area aren’t just in it for the financial returns; they’re also in it for the social returns.

    The Bay Area tech ecosystem has been so successful that startup-related news has become the principal determinant of social status in San Francisco. In other cities, you acquire and flex social status by joining exclusive neighbourhoods or country clubs, or through philanthropic gestures, or even something as simple as what car you drive. In San Francisco, it’s angel investing. Other than founding a successful startup yourself, there’s not much higher-status in the Bay Area than backing founders that go on to build or Stripe…The end result is that the Bay Area has a critical density of people who are willing to offer founders a term sheet for enough investment, and at attractive enough valuations, that it makes sense for the founder to actually accept them. I honestly believe that without this social “subsidy”, a lot of angel investing stops working. If investors were being purely rational, they could only offer something like a $2 million valuation for founders’ first cheques. And if entrepreneurs are smart, they know they can’t accept it; it makes them un-fundable from that day forward.

    The social rewards of angel investing solve an important chicken-and-egg problem in early stage fundraising that financial rewards does not.

    One of the biggest frustrations you face as a founder out fundraising is the refrain: “This sounds really interesting. I love it. Let me know when there are a bunch of other people investing, and then I’ll invest too.” From far away, it’s easy to label this behaviour as cowardly investing. But it happens for a reason…The social returns to angel investing resolve our chicken/​​​​egg problem: they turn angel investing into a kind of “race to be first” that is much more aligned with the founder, and more conducive to breaking inertia and completing deals. The founder wants you to move first, and so do you.

    The social returns to angel investing have a strong geographical network effect, because they require a threshold density in order to kick in.

    …If you can assemble enough early stage investors together, it should conceptually become self-sustaining. Once you have that sufficient density of people who care about the social return to angel investing, and you establish a genuine “early stage capital market” that is subsidized in part by the social and emotional job that it’s doing for its angel members, you create something really special. You get the rare conditions where capital is available for founders at high enough valuations, with no strings attached, and by investors who are evaluating them “the right way”, that you actually sustain a scene that produces startups in sufficient numbers to generate those few unlikely mega-winners that replenish angels’ bank accounts and keep the cycle going.



  6. ⁠, Andreas Pavlogiannis, Josef Tkadlec, Krishnendu Chatterjee, Martin A. Nowak (2018-06-14):

    [] Because of the intrinsic randomness of the evolutionary process, a mutant with a fitness advantage has some chance to be selected but no certainty. Any experiment that searches for advantageous mutants will lose many of them due to random drift. It is therefore of great interest to find population structures that improve the odds of advantageous mutants. Such structures are called amplifiers of : they increase the probability that advantageous mutants are selected. Arbitrarily strong amplifiers guarantee the selection of advantageous mutants, even for very small fitness advantage. Despite intensive research over the past decade, arbitrarily strong amplifiers have remained rare. Here we show how to construct a large variety of them. Our amplifiers are so simple that they could be useful in biotechnology, when optimizing biological molecules, or as a diagnostic tool, when searching for faster dividing cells or viruses. They could also occur in natural population structures.

    In the evolutionary process, mutation generates new variants, while selection chooses between mutants that have different reproductive rates. Any new mutant is initially present at very low frequency and can easily be eliminated by ⁠. The probability that the lineage of a new mutant eventually takes over the entire population is called the ⁠. It is a key quantity of evolutionary dynamics and characterizes the rate of evolution.

    …In this work we resolve several open questions regarding strong amplification under uniform and temperature initialization. First, we show that there exists a vast variety of graphs with self-loops and weighted edges that are arbitrarily strong amplifiers for both uniform and temperature initialization. Moreover, many of those strong amplifiers are structurally simple, therefore they might be realizable in natural or laboratory setting. Second, we show that both self-loops and weighted edges are key features of strong amplification. Namely, we show that without either self-loops or weighted edges, no graph is a strong amplifier under temperature initialization, and no simple graph is a strong amplifier under uniform initialization.

    …In general, the probability depends not only on the graph, but also on the initial placement of the invading mutants…For a wide class of population structures17, which include symmetric ones28, the fixation probability is the same as for the well-mixed population.

    … A population structure is an arbitrarily strong amplifier (for brevity hereafter also called “strong amplifier”) if it ensures a fixation probability arbitrarily close to one for any advantageous mutant, r > 1. Strong amplifiers can only exist in the limit of large population size.

    Numerical studies30 suggest that for spontaneously arising mutants and small population size, many unweighted graphs amplify for some values of r. But for a large population size, randomly constructed, unweighted graphs do not amplify31. Moreover, proven amplifiers for all values of r are rare. For spontaneously arising mutants (uniform initialization): (1) the Star has fixation probability of ~1 − 1⁄r2 in the limit of large N, and is thus an amplifier17, 32, 33; (2) the Superstar (introduced in ref. 17, see also ref. 34) and the Incubator (introduced in refs. 35, 36), which are graphs with unbounded degree, are strong amplifiers.

    …In this work we resolve several open questions regarding strong amplification under uniform and temperature initialization. First, we show that there exists a vast variety of graphs with self-loops and weighted edges that are arbitrarily strong amplifiers for both uniform and temperature initialization. Moreover, many of those strong amplifiers are structurally simple, therefore they might be realizable in natural or laboratory setting. Second, we show that both self-loops and weighted edges are key features of strong amplification. Namely, we show that without either self-loops or weighted edges, no graph is a strong amplifier under temperature initialization, and no simple graph is a strong amplifier under uniform initialization.

    Figure 1: Evolutionary dynamics in structured populations. Residents (yellow) and mutants (purple) differ in their reproductive rate. (a) A single mutant appears. The lineage of the mutant becomes extinct or reaches fixation. The probability that the mutant takes over the population is called “fixation probability”. (b) The classical, well-mixed population is described by a complete graph with self-loops. (Self-loops are not shown here.) (c) Isothermal structures do not change the fixation probability compared to the well-mixed population. (d) The Star is an amplifier for uniform initialization. (e) A self-loop means the offspring can replace the parent. Self-loops are a mathematical tool to assign different reproduction rates to different places. (f) The Superstar, which has unbounded degree in the limit of large population size, is a strong amplifier for uniform initialization. Its edges (shown as arrows) are directed which means that the connections are one-way.
    Figure 4: Infinite variety of strong amplifiers. Many topologies can be turned into arbitrarily strong amplifiers (Wheel (a), Triangular grid (b), Concentric circles (c), and Tree (d)). Each graph is partitioned into hub (orange) and branches (blue). The weights can be then assigned to the edges so that we obtain arbitrarily strong amplifiers. Thick edges receive large weights, whereas thin edges receive small (or zero) weights

    …Intuitively, the weight assignment creates a sense of global flow in the branches, directed toward the hub. This guarantees that the first 2 steps happen with high probability. For the third step, we show that once the mutants fixate in the hub, they are extremely likely to resist all resident invasion attempts and instead they will invade and take over the branches one by one thereby fixating on the whole graph. For more detailed description, see “Methods” section “Construction of strong amplifiers”.

    Necessary conditions for amplification: Our main result shows that a large variety of population structures can provide strong amplification. A natural follow-up question concerns the features of population structures under which amplification can emerge. We complement our main result by proving that both weights and self-loops are essential for strong amplification. Thus, we establish a strong dichotomy. Without either weights or self-loops, no graph can be a strong amplifier under temperature initialization, and no simple graph can be a strong amplifier under uniform initialization. On the other hand, if we allow both weights and self-loops, strong amplification is ubiquitous.

    …Some naturally occurring population structures could be amplifiers of natural selection. For example, the germinal centers of the immune system might constitute amplifiers for the affinity maturation process of adaptive immunity46. Habitats of animals that are divided into multiple islands with a central breeding location could potentially also act as amplifiers of selection. Our theory helps to identify those structures in natural settings.


  8. ⁠, Jonas Richner (2020-07-22):

    …This Flash game is called Canabalt. A businessman crashes out of a window and starts running to escape the destruction of his city. Canabalt sparked the entire endless runner genre of gameplay, which is now one of the most popular genres on mobile. The game has since been included in the New York Museum of Modern Art, alongside Pac-Man and Tetris. Escape room games, now a popular genre, originally came from Flash games. They even made the jump into real life, with many physical escape rooms all over the world. There were many more Flash games. Millions more. Played billions of times on thousands of different gaming websites. It was creative chaos. Flash games were the gateway for many developers in the games industry, and served as an experimental playground for distilling games down to their most pure and engaging elements. The end-of-life of Flash in December 2020 marks the end of one of the most creative periods in the history of gaming. It all started in 1996, when the Flash player was first released. Originally it was intended for Web graphics and animations, but when it got its own programming language in 2000, developers started to use it to make games. That was the same year we saw the rise of the first automated Flash games website, Newgrounds. Anyone could upload their games and they were published immediately…

    [Followed by timeline of Flash games; >20 testimonials from ex-Flash developers and game industry figures.]

  9. ⁠, Jacob Sweet (2021-03-26):

    By 2009, Harry Hong, a spiky-haired twenty-four-year-old Angeleno, delivered the site’s first certified max-out, and Adam Cornelius, another enthusiast and a filmmaker, began working on a documentary about the remarkable achievement. When Harrison saw the project on Kickstarter, he donated a few hundred dollars to help complete the film, but added a caveat. “You can’t just talk about Harry Hong”, he recalls writing. “You’ve got to talk about Jonas Neubauer. You’ve got to talk about Thor Aackerlund. You’ve got to get these guys together and have a tournament and see who’s actually the best.”

    Some of the players who gathered for the first ⁠, for all their thousands of hours of practice, were in the dark about basic tactics. Hong was stunned to learn that his strategy of scoring Tetrises by dropping long bars into a left-side gap was suboptimal. Due to piece-flipping mechanics, a right-side gap was superior. Dana Wilcox, one of the highest-scoring players on the leaderboard, discovered that she’d played for 20 years without knowing that the blocks could be spun in either direction.

    …Learning to “hyper-tap” was a priority. Thor had been the first to hyper-tap, but, by 2017, Koryan Nishio, a Japanese programmer in his forties, was the only prominent player using the technique. (“It seemed like a lot of work for a video game”, Vince Clemente, who has co-organized the classic-Tetris tournament since its inception, explained.) To Joseph, though, it was the obvious way to go. To tap quickly, he developed a unique one-handed grip: with his right thumb on the control pad, he flexed his right bicep until his arm shook, pressing down with each tremor, about fifteen times per second. He turned his thumb into a jackhammer.

    …Jonas quit his job to stream full-time on Twitch—broadcasting an efficient, battle-tested style for amateurs to emulate. When Joseph won the tournament again, in 2019, he inspired more young players. In 2020 alone, 131 players maxed out; between 1990 and 2019, 87 players had maxed out. Kids had killed the Tetris curve.

    These new players see a max-out not as an impossibility, but as a rite of passage. Before even buying the game, most of the rising generation of classic-Tetris players have already watched hours of the best performances, hard-wiring beautiful stacking strategies. As they begin practicing, they often join one of many classic-Tetris servers on Discord, where hundreds of people are online all the time, ready to discuss any aspect of the game. It’s there that they often learn the most common hyper-tapping grip—holding the controller sideways, with the directional pad facing up—and how to properly tense the right arm so that it shakes quickly and consistently. They study the principles of developing a relatively even stack with a built-out left side, and discuss how dropping a pair of tetrominoes in a complementary orientation can reduce the need for a timely T-piece. They can imitate Joseph’s “hyper-tap quick-tap”, in which he sneaks in a left-handed tap among a right-thumb flurry, or watch Jonas’s “Tetris Spin Class” and observe how certain flips can clear a line and make the stack Tetris ready.

    What took Jonas years to figure out takes new players minutes. “You don’t need to experiment for hours trying to figure out what works and what doesn’t”, Jacob Huff, a nineteen-year-old who maxed out last March after playing for two months, said. “You can ask someone in the Discord and they’ll tell you every spin that you can do.” Strategies born on Discord are practiced and scrutinized on Twitch, then put to the test in a growing pool of competitions: Classic Tetris Monthly, Classic Tetris League, Classic Tetris Gauntlet, Classic Tetris Brawl. Thanks to hyper-tapping and more efficient stacking, players build higher and higher, almost refusing to accept any line clearance that’s not a Tetris. To the older generation, the style seems reckless. To newer players, it’s simply the best way to play.

    …By the quarter-final [of the championship], the entire old guard had vanished. The remaining players were all of the YouTube generation, with many explicitly crediting its algorithm for introducing them to classic Tetris.



  12. ⁠, Brendan I. Koerner (2021-05-27):

    When Pinel looked into the discourse around ball performance, he found that most everyone believed that all that mattered was the quality of coverstock—that is, the exterior layer of a ball that is visible to the naked eye. Coverstocks are studded with microscopic spikes, the roughness of which is measured by the average distance from each spike’s peak to valley—a metric known as Ra. The higher a ball’s Ra, the more friction it can create with the lane and thus the greater the potential that it will hook well under the right circumstances. The hardness of the material that underlies the spikes is also an important factor. In the early 1970s, several pros had enjoyed great success by soaking their balls in methyl ethyl ketone, a flammable solvent that softened the coverstocks. (The balls became so gelatinous, in fact, that a bowler could indent the surface with a fingernail.) These softer balls gripped the lane much better than their harder counterparts, and so they tended not to skid unpredictably when encountering patches of oil used to dress the wooden boards. The use of methyl ethyl ketone had increased scores so much that rules were put in place mandating a degree of coverstock hardness as measured by a device known as a Shore durometer.

    Pinel thought that too much attention was being paid to coverstocks and not nearly enough to what was inside the ball…Pinel used the shop’s drill and off-the-shelf components to alter balls. He’d pock them with deep holes that he’d then fill with dense wads of barium, a soft metal. “So I’d drill a hole, fill it with either dense or light stuff, and plug it to the top”, he says. “And I started playing around with that, and I started to see some differences in motion.”

    Never lacking confidence, Pinel contacted several ball manufacturers in 1973 and proposed a deal: If they would sign a nondisclosure agreement, he’d brief them on his experimental results and help them design balls that would allow amateurs and pros alike to increase their strike rates. Company executives responded that they were willing to listen to Pinel’s ideas, but he was the one who would have to sign a release affirming that nothing he said was confidential. Miffed by what he saw as attempts to steal his ideas, Pinel veered away from a career in ball design.

    …The AMF Sumo, the smash-hit ball that would earn Pinel his kanji pendant, was released in 1992. This time, Pinel opted for a core that bears a passing resemblance to the video game character Q*Bert, albeit with a disc at the base in lieu of feet. The ball came out right as new regulations called for more oil to be poured on lanes, a change that decreased friction; this sapped shots of spin and power. The extra oil was no match for the Sumo, however, because Pinel’s core caused it to slice hard across the boards near the pins. The ball would eventually sell well enough to make Pinel a modestly wealthy man…Pinel says the size of his royalty eventually became a problem for AMF, and the company terminated his contract in 1995. He was barely out of work a week before he was hired by Faball…The ball that contained this revamped core, the Hammer 3D Offset, would become Pinel’s signature achievement. “That ball sold like hotcakes for 3 years, where the average life span of a ball was about 6 months”, says Del Warren, a former ball designer who now works as a coach in Florida. “They literally couldn’t build enough of them.” In addition to flaring like few other balls on the market, the 3D Offset was idiot-proof: The core was designed in such a way that it would be hard for a pro shop to muck up its action by drilling a customer’s finger holes incorrectly, an innovation that made bowlers less nervous about plunking down $398$2001996 for a ball…Pinel was delighted by the 3D Offset’s success not just because it affirmed his beliefs about the importance of asymmetry but also because it inflicted pain on his former employer. “AMF had been doing $24$121996 million a year; Hammer had been doing $2$11996 million”, he told me. “When we came out with the 3D Offset, Hammer did $24$121996 million a year and AMF did $2$11996 million. Not that I enjoyed that at all.” AMF would file for bankruptcy 4 years later.

    …Pinel was still trying to maximize flare potential in his designs, an effort that was arguably becoming outmoded. A new generation of pro bowlers, both stronger and more technically sophisticated than their predecessors, have achieved unprecedented amounts of spin on their balls—sometimes as much as 600 revolutions per minute for those who opt for the increasingly popular 2-handed throwing technique. Such bowlers don’t need as much hook assistance as in days gone by, so they’re using more stable balls—a strategic trend that may be having a trickle-down effect on the league bowlers who worship the sport’s stars. In our conversations, Pinel never displayed any hint that he was worried about the future of his cores.



  15. ⁠, James Surowiecki () (2014-11-03):

    [Discussion of the creation of modern sports training: professional athletes, even NBA stars, typically did not ‘train’. Practice was about getting into shape and working with teammates, if even that much—one simply took one’s skills for granted. Coaches focused on strategy, not coaching.

    A harbinger of the professionalization of professional athletes was basketball player Kermit Washington, on the verge of washing out of the NBA early on until he swallowed his pride and began tutoring with coach Pete Newell, who drilled Kermit on the basics repeatedly. Kermit eventually became an All-Star player and influenced other NBA players to engage in coaching and deliberate practice to improve their fundamentals. The modern paradigm is a ruthless quest for perfection in every dimension, quantified, and applying the latest science and technology to eek out even the slightest fraction of a second improvement; athletes are projects, with many different specialists examining them constantly for potential improvements, and as importantly, when not to practice lest they be injured.

    And the results speak for themselves—performance has never been higher, the impossible is now done routinely by many professionals, this continuous improvement trend has spread to other domains too, including chess, classical music, business. Equally striking are domains which don’t see trends like this, particular American education.]

    “You need to have the best PhDs onboard as well”, McClusky says. This technological and analytical arms race is producing the best athletes in history.

    The arms race centers on an obsessive scrutiny of every aspect of training and performance. Trainers today emphasize sports-specific training over generalized conditioning: if you’re a baseball player, you work on rotational power; if you’re a sprinter, on straight-line explosive power. All sorts of tools have been developed to improve vision, reaction time, and the like. The Dynavision D2 machine is a large board filled with flashing lights, which ballplayers have to slap while reading letters and math equations that the board displays. Football players use Nike’s Vapor Strobe goggles, which periodically cloud for tenth-of-a-second intervals, in order to train their eyes to focus even in the middle of chaos. Training is also increasingly personalized. Players are working not just with their own individual conditioning coaches but also with their own individual skills coaches. In non-team sports, such as tennis and golf, coaches were rare until the seventies. Today, tennis players such as Novak Djokovic have not just a single coach but an entire entourage. In team sports, meanwhile, there’s been a proliferation of gurus. George Whitfield has built a career as a “quarterback whisperer”, turning college quarterbacks into NFL-ready prospects. Ron Wolforth, a pitching coach, is known for resurrecting pitchers’ careers—he recently transformed the Oakland A’s Scott Kazmir from a has-been into an All-Star by revamping his mechanics and motion. Then there’s the increasing use of biometric sensors, equipped with heart-rate monitors, G.P.S., and gyroscopes, to measure not just performance (how fast a player is accelerating or cutting) but also fatigue levels. And since many studies show that getting more sleep leads to better performance, teams are now worrying about that, too. The N.B.A.’s Dallas Mavericks have equipped players with Readiband monitors to measure how much, and how well, they’re sleeping.

    All this effort may sound a bit nuts. But it’s how you end up with someone like Chris Hoy, the British cyclist who won two gold medals at the London Olympics in 2012, trailed by a team of scientists, nutritionists, and engineers. Hoy ate a carefully designed diet of five thousand calories a day. His daily workouts—two hours of lifting in the morning, three hours in the velodrome in the afternoon, and an easy one-hour recovery ride in the evening—had been crafted to maximize both his explosive power and his endurance. He had practiced in wind tunnels at the University of Southampton. He had worn biofeedback sensors that delivered exact data to his trainers about how his body was responding to practice. The eighty-thousand-dollar carbon-fibre bike he rode helped, too. Hoy was the ultimate product of an elaborate and finely tuned system designed to create the best cyclist possible. And—since his competitors weren’t slacking, either—he still won by only a fraction of a second.

  16. ⁠, Atul Gawande (New Yorker) (2011-10-26):

    [Meditation by doctor interested in medical improvement/​​​​progress (elsewhere, checklists). In tennis, he had improved his performance enormously after just minutes of coaching from a young man who pointed out his errors. Coaches are used in many areas and often spot problems that highly-competent trained professionals continue to make. A good coach is emotionally supportive, careful, speaks with credibility so they are not reflexively dismissed, brings an independent eye to highlight blind spots, and always finds a way they can push themselves to improve and deliberately practice.

    Gawande, having noticed his surgery success rates plateaued, considers a ‘medical coach’. Doctors are intensively taught up until they become full-fledged doctors, at which point they are cut loose to act as little gods in their domains, with no supervision. Yet, they are almost surely not perfect, and their skills may degrade over time. In domains far less important, like entertainment (arts/​​​​athletics), no individual believes they are perfect and they use personal coaches to constantly critique themselves, spot errors that untrained eyes would not, and strive for improvement. Why don’t we do the same thing in important things like surgeries? Why not coaches for doctors? Does the mystique of doctors intimidate themselves (and patients) away from acknowledging error and fallibility and improving?

    Gawande talks a former medical professor into coaching him. Gawande, while proud of his surgical technique, is surprised how many flaws his coach notes, and embarrassed; he had become used to working on his own, with no accountability or external critique. Other doctors made fun of the idea of coaching (coaching for thee, not for me). But he worked on his errors, and feels positive about his improvements and the possibility of breaking out of his plateau.]


  18. ⁠, Péter Poczai, Neil Bell, Jaakko Hyvönen (2014-01-21):

    Contemporary science thrives on collaborative networks, but these can also be found elsewhere in the history of science in unexpected places. When Mendel turned his attention to inheritance in peas he was not an isolated monk, but rather the latest in a line of Moravian researchers and agriculturalists who had been thinking about inheritance for half a century. Many of the principles of inheritance had already been sketched out by Imre Festetics, a Hungarian sheep breeder active in Brno. Festetics, however, was ultimately hindered by the complex nature of his study traits, aspects of wool quality that we now know to be polygenic. Whether or not Mendel was aware of Festetics’s ideas, both men were products of the same vibrant milieu in 19th-century Moravia that combined theory and agricultural practice to eventually uncover the rules of inheritance.




  22. ARPA








  30. Parasocial


  32. ⁠, Richard Hamming (1986-03-07):

    [Transcript of a talk by mathematician and manager about what he had learned about computers and how to do effective research (republished in expanded form as Art of Doing Science and Engineering: Learning to Learn; 1995 video). It is one of the most famous and most-quoted such discussions ever.]

    At a seminar in the Bell Communications Research Colloquia Series, Dr. Richard W. Hamming, a Professor at the Naval Postgraduate School in Monterey, California and a retired Bell Labs scientist, gave a very interesting and stimulating talk, ‘You and Your Research’ to an overflow audience of some 200 Bellcore staff members and visitors at the Morris Research and Engineering Center on March 7, 1986. This talk centered on Hamming’s observations and research on the question “Why do so few scientists make substantial contributions and so many are forgotten in the long run?” From his more than 40 years of experience, 30 of which were at Bell Laboratories, he has made a number of direct observations, asked very pointed questions of scientists about what, how, and why they did things, studied the lives of great scientists and great contributions, and has done introspection and studied theories of creativity. The talk is about what he has learned in terms of the properties of the individual scientists, their abilities, traits, working habits, attitudes, and philosophy.



  35. on-really-trying




  39. Bakewell#social-contagion

  40. Timing#try-try-again-but-less-less

  41. Backstop#internet-community-design

  42. Scaling-hypothesis

  43. Copyright-deadweight

  44. Local-optima

  45. Competence