3 ms·

Great post. I wouldn't be so defeatist about technological solutions though. I can think of a few technologies that we already have which have allowed intimate,

Great post. I wouldn't be so defeatist about technological solutions though. I can think of a few technologies that we already have which have allowed intimate, important and meaningful discussions to scale (imperfectly) in our past. 1. Parliamentary proceducers. This is a tool that allows debate to scale up to a few hundred/thousand people. It provides rules for how topics can be chosen for discussion, altered, debated and how to reach consensus during a discussion. It's a social technology that is not easy to apply, not obvious, but used by dozens of countries and I've no clue how many large organizations. 2. The judicial and academic systems. Again, not flawless, but these ones show that we can have asynchronous debates over long periods of time where people reference and use old cases and studies. There may be other mechanisms which humans have used to structure conversation, facilitate debate and allow for important expression. All of these require training and only work when participants agree to certain rules. To my knowledge none of these apply to online media. But, who knows, maybe it will take a few hundred years for us to discover some social technology that does. The examples I bring were not obvious to society and took similar timescales to develop.

high_5 3y ago
Both 1 and 2 provide good layout for important and meaningful, but not intimate discussions. Most social conversations are not necessarily obliged towards a resolution of a problem as in these two cases. In intimate discussions it's most of the time good enough to be heard.
dredmorbius 3y ago
Those are two excellent examples, as they're certainly technical, but not based on advanced information technologies, which is a common bias in HN discussions. I've done some informal looking into governmental structures, with one key entity often being some sort of central cabinet, central committee, politburo, or privy council, which can be found in a wide range of governmental systems: representative democracies, parliamentary, Communist, and monarchical. In almost all cases, that group tends to be on the order of 5--9 or so individuals. This would mean that, say, the US Cabinet (28 members) would have a much smaller core group, with the most powerful positions generally being considered as the Secretaries of State, Treasury, and Defence. Along with the National Security Advisor and perhaps the Attorney General, this would give a core counsel of six members, including the President. Parliaments virtually always operate in terms of both committees and parties, with each of these being further subdivided. US House and Senate committees still tend to be large, roughly 25--50 members. Taking the US Senate Appropriations Subcommittee for Defence (one of twelve Appropriations subcommittes), there are 17 members, divided amongst the majority (9) and minority (8) parties. Taken by party, this gets us to my five-to-nine member core working group. There's typically further breakout by ad hoc working groups, and as I understand there are unelected staff (that is, subcommittee participants who are not themselves elected members of Congress) as well. There's little actual work which occurs in full sessions of a legislature. Votes (based on Committee work), speeches (largely for public consumption, and often to an empty chamber), ceremonial proceedings (swearing in, State of the Union), and with exceptions such as impeachment proceedings or the largely-but-not-entirely-it-seems perfunctory functions of certifying national election results. What parliamentary procedures do provide for is smooth function of subdivisions of full legislatures (e.g., committees and subcommittees), as well as rules for overall operation of the largely pro forma full-chamber sessions. That said, procedures are key in those functions. The Judicial system is interesting in that courts often operate highly autonomous, with individual judges having extreme discretion over the operations of their courtrooms, and courts collectively of dockets. A US trial court consists of a single judge. If held before a jury, there are usually 12 jurors in civil and criminal cases (fewer in some state courts), though jurors don't actively participate in the trial proceedings themselves, but rather observe the prosecution / plaintiff and defence, as well as judge's orders and rulings. After completion of the trial phase, the jury members deliberate amongst themselves to reach a verdict. Witnesses are fully independent (and are often excluded from hearing one anothers' testimony), legal teams are also typically small in all but the very largest cases. Other countries have various different systems, often with civil or criminal cases being tried before a panel of judges (frequently three). US appeals trials operate similarly, though in significant cases an en banc hearing with all the judges in the district hearing the case (six to twelve, I believe, depending on the circuit). Appeals courts and Supreme courts, whether at the state, Federal Circuit, or Supreme Court levels are the only overview of individual court judges and judgements in the US. Which is to say, again, that the structure is highly independent and autonomous and based on small groups. Academia is another interesting situation, again with relatively high autonomy amongst tenured faculty, though that's been decreasing. Academia is divided into disciplines, universities, colleges, and departments. Typical department sizes again tend to fall into the 15--50 full-time permanent members (there may be many more teaching assistants, lecturers, and other part-time or contingent positions). You didn't mention business organisation, though that would be another useful model. I believe the terms are M-Type and D-Type for monolithic and divisional corporate structures, respectively, with the latter emerging at firms such as DuPont and General Motors (particularly under Alfred P. Sloan, Jr., at GM), or terms to that effect. Again, the goal is to reduce dependencies between organisational divisions whilst also achieving efficiencies of scale. (Charles Perrow devoted much of his academic life to studying organisations, and wrote survey text on the topic, Complex Organizations, which discusses much of the prior work in the field.) All of which again generally show that communications scales poorly. Where it does scale, it almost always does so by hugely simplifying the content communicated. Markets, where prices substitute for a whole host of other qualities, would be an example of this. (That prices communicate complex qualitative differences and nuances poorly is another well-known challenge. Complexity will out.) Information automation does expand processing capability, but that tends to end up in a Red Queen's race, and often seems to consume free budget in additional processing, effectively a form of economic rent, rather than delivering improved profits, a/k/a the Solow Productivity Paradox.

high_5 3y ago

Both 1 and 2 provide good layout for important and meaningful, but not intimate discussions. Most social conversations are not necessarily obliged towards a resolution of a problem as in these two cases. In intimate discussions it's most of the time good enough to be heard.

dredmorbius 3y ago

Those are two excellent examples, as they're certainly technical, but not based on advanced information technologies, which is a common bias in HN discussions. I've done some informal looking into governmental structures, with one key entity often being some sort of central cabinet, central committee, politburo, or privy council, which can be found in a wide range of governmental systems: representative democracies, parliamentary, Communist, and monarchical. In almost all cases, that group tends to be on the order of 5--9 or so individuals. This would mean that, say, the US Cabinet (28 members) would have a much smaller core group, with the most powerful positions generally being considered as the Secretaries of State, Treasury, and Defence. Along with the National Security Advisor and perhaps the Attorney General, this would give a core counsel of six members, including the President. Parliaments virtually always operate in terms of both committees and parties, with each of these being further subdivided. US House and Senate committees still tend to be large, roughly 25--50 members. Taking the US Senate Appropriations Subcommittee for Defence (one of twelve Appropriations subcommittes), there are 17 members, divided amongst the majority (9) and minority (8) parties. Taken by party, this gets us to my five-to-nine member core working group. There's typically further breakout by ad hoc working groups, and as I understand there are unelected staff (that is, subcommittee participants who are not themselves elected members of Congress) as well. There's little actual work which occurs in full sessions of a legislature. Votes (based on Committee work), speeches (largely for public consumption, and often to an empty chamber), ceremonial proceedings (swearing in, State of the Union), and with exceptions such as impeachment proceedings or the largely-but-not-entirely-it-seems perfunctory functions of certifying national election results. What parliamentary procedures do provide for is smooth function of subdivisions of full legislatures (e.g., committees and subcommittees), as well as rules for overall operation of the largely pro forma full-chamber sessions. That said, procedures are key in those functions. The Judicial system is interesting in that courts often operate highly autonomous, with individual judges having extreme discretion over the operations of their courtrooms, and courts collectively of dockets. A US trial court consists of a single judge. If held before a jury, there are usually 12 jurors in civil and criminal cases (fewer in some state courts), though jurors don't actively participate in the trial proceedings themselves, but rather observe the prosecution / plaintiff and defence, as well as judge's orders and rulings. After completion of the trial phase, the jury members deliberate amongst themselves to reach a verdict. Witnesses are fully independent (and are often excluded from hearing one anothers' testimony), legal teams are also typically small in all but the very largest cases. Other countries have various different systems, often with civil or criminal cases being tried before a panel of judges (frequently three). US appeals trials operate similarly, though in significant cases an en banc hearing with all the judges in the district hearing the case (six to twelve, I believe, depending on the circuit). Appeals courts and Supreme courts, whether at the state, Federal Circuit, or Supreme Court levels are the only overview of individual court judges and judgements in the US. Which is to say, again, that the structure is highly independent and autonomous and based on small groups. Academia is another interesting situation, again with relatively high autonomy amongst tenured faculty, though that's been decreasing. Academia is divided into disciplines, universities, colleges, and departments. Typical department sizes again tend to fall into the 15--50 full-time permanent members (there may be many more teaching assistants, lecturers, and other part-time or contingent positions). You didn't mention business organisation, though that would be another useful model. I believe the terms are M-Type and D-Type for monolithic and divisional corporate structures, respectively, with the latter emerging at firms such as DuPont and General Motors (particularly under Alfred P. Sloan, Jr., at GM), or terms to that effect. Again, the goal is to reduce dependencies between organisational divisions whilst also achieving efficiencies of scale. (Charles Perrow devoted much of his academic life to studying organisations, and wrote survey text on the topic, Complex Organizations, which discusses much of the prior work in the field.) All of which again generally show that communications scales poorly. Where it does scale, it almost always does so by hugely simplifying the content communicated. Markets, where prices substitute for a whole host of other qualities, would be an example of this. (That prices communicate complex qualitative differences and nuances poorly is another well-known challenge. Complexity will out.) Information automation does expand processing capability, but that tends to end up in a Red Queen's race, and often seems to consume free budget in additional processing, effectively a form of economic rent, rather than delivering improved profits, a/k/a the Solow Productivity Paradox.