4 ms·

What the hell does he even mean by Tech Monopolies? People today are just making up words.

jeffreyrogers 6y ago
There are literally antitrust suits against Facebook and Google currently, so at least some part of the Justice department believes these tech companies are monopolies, as does a sizeable, bipartisan part of Congress.
- franklampard 6y ago
  Okay, but I am sure the suits are not against web infra parts of those businesses, right? Because in this context he is taking about web infrastructure, and specially AWS.
lukifer 6y ago
Networked services intrinsically reward monopolization/centralization, due to Metcalfe's Law [0]: users are more concerned with going where other users are, than the merits of the service itself. (A curious exception: Uber and Lyft, presumably because drivers can easily use both simultaneously.) Facebook, for example, does not represent a monopoly, defined as a social media company generally; but they arguably have ~99% market share within their social niche (with some FOSS clones as a rounding error, and Twitter/Instagram/etc occupying different niches). Contrast with Netflix, which is more disruptable without social network effects, as we've seen with the proliferation of streaming services in the last 5 years. I don't know what the threshold should be to regulate social media as utilities (mandating transparency, interoperability, and due process), but as such platforms become our de-facto political and conversational commons, I don't think we should allow those spaces to be entirely controlled by private capital, let alone manipulated by Blind Idiot God algorithms, which happily hijack our limbic systems for a quick CPM buck. [0] https://en.wikipedia.org/wiki/Metcalfe%27s_law https://en.wikipedia.org/wiki/Metcalfe%27s_law
- dredmorbius 6y ago
  Ben Tilly (@btilly here) and Andrew Odlyzko's "A refutation of Metcalfe's Law and a better estimate for the value of networks and network interconnections" gives a better estimator, n * log(n). But otherwise yes. http://www.dtc.umn.edu/~odlyzko/doc/metcalfe.pdf http://www.dtc.umn.edu/~odlyzko/doc/metcalfe.pdf My view is that there is also a costant-across-n (thouggh variable with t) cost function, so: v = n * log(n) - k[t](n) Where cost exceeds value, growth stops. Networks grow until log(n) = k[t](n) n == members t == time

There are literally antitrust suits against Facebook and Google currently, so at least some part of the Justice department believes these tech companies are monopolies, as does a sizeable, bipartisan part of Congress.

franklampard 6y ago
Okay, but I am sure the suits are not against web infra parts of those businesses, right? Because in this context he is taking about web infrastructure, and specially AWS.

lukifer 6y ago

Networked services intrinsically reward monopolization/centralization, due to Metcalfe's Law [0]: users are more concerned with going where other users are, than the merits of the service itself. (A curious exception: Uber and Lyft, presumably because drivers can easily use both simultaneously.) Facebook, for example, does not represent a monopoly, defined as a social media company generally; but they arguably have ~99% market share within their social niche (with some FOSS clones as a rounding error, and Twitter/Instagram/etc occupying different niches). Contrast with Netflix, which is more disruptable without social network effects, as we've seen with the proliferation of streaming services in the last 5 years. I don't know what the threshold should be to regulate social media as utilities (mandating transparency, interoperability, and due process), but as such platforms become our de-facto political and conversational commons, I don't think we should allow those spaces to be entirely controlled by private capital, let alone manipulated by Blind Idiot God algorithms, which happily hijack our limbic systems for a quick CPM buck. [0] https://en.wikipedia.org/wiki/Metcalfe%27s_law https://en.wikipedia.org/wiki/Metcalfe%27s_law

dredmorbius 6y ago
Ben Tilly (@btilly here) and Andrew Odlyzko's "A refutation of Metcalfe's Law and a better estimate for the value of networks and network interconnections" gives a better estimator, n * log(n). But otherwise yes. http://www.dtc.umn.edu/~odlyzko/doc/metcalfe.pdf http://www.dtc.umn.edu/~odlyzko/doc/metcalfe.pdf My view is that there is also a costant-across-n (thouggh variable with t) cost function, so: v = n * log(n) - k[t](n) Where cost exceeds value, growth stops. Networks grow until log(n) = k[t](n) n == members t == time