Until 1995, the internet was only available for specialists. Yes, it only took 5 years then to the first dot com bubble.
Because the business model of telecom companies changed.
You see --
Ca. 1995 in Poland the situation was that you had local ISPs which could afford to lay a city-wide fiber, but it could not afford to lay long distance fiber. So --
They were renting long-distance data transmission circuits from the telecom company (TPSA) which already had inter-city fiber. Problem was, TPSA was not in the data transmission business, but in the phone business. And, remember that voice is a low-bandwidth thing: a 2Mbps channel can carry 30 phone connections simultaneously. So TPSA had a nationwide fiber network running at 34Mbps which was more than enough for the voice business --
And when a customer (ISP, bank, whatever) came to TPSA and said that they need a dedicated 2Mbps circuit between Krakow and Kielce, then this circuit represented 1/17th of the pipe available on that route (and loss of 30 slots for very profitable long-distance phone calls). Worse yet, a virtual circuit between Krakow and Gdansk would be actually going Krakow -> Kielce -> Warsaw -> Elblag -> Gdansk, eating 1/17th of the pipe on each leg --
So TPSA would provide the inter-city virtual circuits, but would charge very high prices for them. That, of course, made the internet very expensive for the end users, as the end-user was essentially paying ISP's overhead on top of TPSA's racket price, but --
The rising demand for virtual circuits caused TPSA to replace the transmission equipment, going from 34Mbps to 622Mbps on the same fiber --
At this point, someone at TPSA saw the light and realized that since TPSA now has a ton of long-haul bandwidth available, instead of selling virtual circuits to middlemen they can simply cut them out, first by offering IP transit from existing phone exchanges (removing the need for virtual circuits), and then by installing dial-in modems on these exchages -- and charging customers by the minute --
And that of course caused the data traffic to skyrocket dramatically shifting priorities, so TPSA sort of moved from the phone business to the data transmission business, so now their main product actually is a data service with the phone service being an add-on.
So back to the subject --
The Internet revolution happened because the demand for data services on voice networks was creeping up, overloading the networks, which forced a data-centered network rebuild, but once that happened the data transmission cost has dropped dramatically, and that created a feedback loop drawing in more customers causing further drop in prices --
And the situation with space access is that we seem to be departing from the $10K/kg price point and moving down and that makes space access affordable for new customers --
And that should create a feedback loop which should further drive the costs down.
Please observe that this analysis ignores the business model of the launch customer. It assumes that, like with the internet, the increased access to the fundamental service (data transmission / space launch) will create new business models, even those that cannot be even envisaged at present --
I mean, would you imagine in 1995 that there would be people in 2015 who make their living posting prank videos on youtube? Right. So why do you assume that a viable space-based business model does not exist at lower launch prices?
Another lesson from my telecom days --
Experts usually overestimate the technology level needed to create a successful business. For example, ca. 2000 it was a common wisdom that video transmission over Internet (specifically TCP) will be problematic because of network congestion, and a lot of academic papers were written on how to deal with the problem. Then came youtube. Their solution? Just purchase big enough pipes running to their servers...
