Saturday, August 20, 2016

More Competition Should Help End Need for Net Neutrality Rules

The T-Mobile US “Binge On” program neatly illustrates the tough issues inherent in network neutrality rules. The rules themselves allow for network management.

And yet it is hard to distinguish between network management and violations of the rules.

Also, network management is designed to preserve user experience under conditions of high network load. But that is among the aims of programs such as Binge On, which
reduce entertainment video to standard definition to enable unlimited use, while not crashing the network.

One can argue that networks now have so much bandwidth that management is not required, or that the way to avoid network management is simply to invest more capital in networks. Certainly, that approach always is taken, to some degree.

Still, few networks exist in a state of permanent abundance, which would eliminate the need for management.

From time to time, any particular network can be lightly loaded. But success causes load. So unless a network attracts few customers, it eventually becomes susceptible to congestion.

Net neutrality rules were seen as a way to ensure consumer choice (no blocking or throttling of lawful apps). But the rules also prevent consumer choice (no lawful content delivery features, for example). Some also view sponsored apps or data as violations of the rules.

Reasonable people will disagree about those matters.

Still, with the passage of time, we are going to see additional examples of instances where net neutrality and network management, or net neutrality and consumer benefit, will clash.

Many argue that robust competition “fixes” the problem of potential antitrust behavior or exercise of market power. Hopefully that will soon come to be the case for consumer Internet access, and we will not need net neutrality rules, which increasingly will have negative impact on efforts to innovate.

More competition, and more supply, should alleviate the need for the rules.

No comments:

Costs of Creating Machine Learning Models is Up Sharply

With the caveat that we must be careful about making linear extrapolations into the future, training costs of state-of-the-art AI models hav...