Saturday, April 2, 2016

Is "Artificial Intelligence" S Curve Going to Affect Telecom?

One recurring pattern in modern technology and for most products, including communications technology, is its nonlinear character. Many call that an “S curve.” The principle also is highly similar to the product life cycle.

Experientially, the S curve is important because it plays a part in human expectations about disruptive new technology. We often expect more change early on, do not see it, and conclude that the trend will not happen. Instead, there is a gestation period, and then an inflection point.

After that, the new trend takes hold rapidly, producing far more change than one might have expected if the process of technology adoption were linear.

That might be worth keeping in mind over the next decade as we make more advances in machine learning, often popularly known as "artificial intelligence." In fact, one might argue that machine intelligence is part of the earlier hype about "big data analytics." The whole point of machine learning is that programs can autonomously "learn" to behave differently, based on new data.

Driverless cars perhaps provide one obvious example of applied big data analytics and machine learning.

Artificial intelligence might be increasingly important for some parts of the telecom business in the future as well.

Much "value" in the sales process traditionally comes from "expert advice" supplied by sales personnel. What happens if much of the expert advice can be supplied by machines, delivering advice over Internet mechanisms?

How might enterprise sales processes change if expert advice can be obtained from smart software, circumventling human channels?

We have seen glimmers of this in other industries, especially industries whose products are amenable to digital substitutiion. 

“Time and again, we have seen digital disruption fundamentally erode value across many industries including: music sales, video rentals, travel booking, and newspapers,” says the Citi Digital Strategy Team.

“In each of these cases, incumbents either transformed or became marginalized,” a Citi report says.

“Digital disruption in these industries resulted on average in a 44 percent share-shift from physical to digital business models over a 10-year period,” according to Citi’s Digital Strategy team.

Market share shifts gradually (1.6 percent per year) until an inflection point around year four when traditional share declines rapidly accelerate to about six percent per year, Citi says.

To say the global telecom business is “different” now since a wave of worldwide deregulation, privatization, investment and emergence of the Internet is a vast understatement. But consider just a few of the biggest changes.

Pre-1980, most national telecom infrastructures were dominated by a single monopoly provider, often owned by the government. In fact, telecommunications was widely believed to be a natural monopoly.

So prices were high, profit margins were high, innovation was low, revenue growth very limited, and the size of the market quite stable.

After deregulation and privatization, prices dropped dramatically, investment increased, total revenue increased, profit margins dropped and rates of innovation climbed dramatically.

In 1991, state-owned telcos numbered about 150. By 2008, that number had dropped to about 70, according to the International Telecommunications Union. By 2008, some 125 nations had fully or partially privatized former state-owned telecom companies, lead especially by mobile operations.

At the same time, the emergence of the Internet and mobility reshaped platforms, the way applications are developed, the power and influence wielded within the ecosystem, business models and value drivers.

At the same time, communications capabilities were rapidly extended to those who previously had no access, rather suddenly transforming a market where perhaps half the world’s people could not “make a phone call” to a world where most adults now have such access, after just a few decades.

And while a similar change, allowing everyone Internet access, is only now underway, it is reasonable to expect that challenge also will be solved, more rapidly than anyone originally might have believed possible.

Three Decades of Disruption
1980
2015
Natural monopoly
Oligopoly
High margin
Moderate to low margin
Low to moderate adoption
High adoption
Low innovation
High innovation
Stable markets
Unstable markets
Compete on quality
Compete on price
Fixed network dominates
Mobile network dominates
Tightly integrated apps and network
Open network
Owned app creation
3rd-party app creation
Sell app, use network access
Sell network access (dumb pipe)
Voice business model
Internet access, mobile business model
Similar business models globally
Growing diversity of business models
99.999% uptime
99.9% or “good enough” availability
Few lead apps
Many lead apps
IT adoption: enterprise; SMB; consumer
IT adoption: consumer/SMB to enterprise

The basic idea is that innovations tend to start slow, but hit an inflection point, then grow rapidly, until hitting a mature and then a declining phase.

Most business and technology developments in the telecommunications business, especially those related to applications and products related to the Internet, seem to follow the S curve.


Adoption now tends to occur very fast. Mobile phones and smartphones provide excellent examples, compared to “earlier” technology such as personal computers.

Adoption of social apps such as WhatsApp show the same pattern.



The point is that most trends have less impact initially than you might expect, but then hit an inflection point and have much more impact than you might expect, were change a linear function.

The implication is that many important changes will seem to go for some period of time where the impact is not seen, but then can gain acceptance very quickly.

“Technology does not just change distribution models and service patterns,” a report by Citi says.  
“The definition of financial products themselves may need to be rethought.”

”We’ll probably be the last generation to use the term credit card and debit card,” said John Stumpf, Wells Fargo CEO. “It will probably be debit access and credit access and it will be likely loaded onto a mobile device.”

No comments:

"Tokens" are the New "FLOPS," "MIPS" or "Gbps"

Modern computing has some virtually-universal reference metrics. For Gemini 1.5 and other large language models, tokens are a basic measure...