AI Will Produce Winners and Losers

Though many executives and analysts are trying very hard to figure out which firms benefit most from generative artificial intelligence and AI in general, the prior experience of firms with the internet suggests there also will be losers.

And those losers could come from industries focused on digital and physical products such as “print” media, as our experience with the internet suggests. 

Study Title	Authors	Year	Focus	Key Findings
How the Internet Changed the Market for Print Media	NBER	2019	Impact on print media	Household adoption of broadband significantly reduced print readership and circulation, leading to revenue decline for newspapers.
The Impact of the Internet on Media Industries: An Economic Perspective	Oxford University Press	2008	Economic impact on media	The internet weakens intellectual property protection, making it easier to distribute content illegally and reducing potential revenue.

Similar losses can be noted in retailing as well, with a shift from place-based and physical retail to online retail. 

Study Title	Authors	Publication Year	Key Findings
The Impact of E-Commerce on Retail Employment	Autor, D., Dorn, D., Hanson, G.	2017	Found that increased e-commerce adoption led to job losses in retail sectors most susceptible to online substitution (e.g., electronics).
Omnichannel Retailing and Customer Engagement: A Review of the Literature	S. Verhoef, M. Kannan, P. Bharadwaj	2009	Highlights the importance of omnichannel strategies for retailers to enhance customer engagement and satisfaction in today's digital age.
The Impact of Online Shopping on Brick-and-Mortar Stores	T. Van den Poel, R. Verhoef	2003	One of the earlier studies exploring the potential negative impacts of e-commerce on traditional brick-and-mortar retailers.

Advertising has seen some of the greatest shifts from the internet, though.

Source: Gemini

Put simply, digital now claims up to 82 percent of all U.S. ad placements and revenue. Print has declined from 42 percent to less than three percent. Linear video dropped from 38 percent to 16 percent. Radio dipped from 10 percent to half a percent. 

Channel	1996 (Billions)	1996 (%)	2023 (Billions)	2023 (%)
Print (Newspapers, Magazines)	80.0	42.1%	10.0	2.7%
Linear Video (TV Broadcast, Cable)	72.0	37.9%	60.0	16.2%
Network Radio	10.0	5.3%	2.0	0.5%
Other (Radio Spots, Out-of-Home)	28.0	14.7%	18.0	4.9%
Digital Ads (Search, Social Media, Display)	-	-	300.0	81.7%

It might be reasonable to expect the content industries, advertising and retailing will again be among the industries to see early AI disruptions. 

Financial services might also be included on the list of industries that saw early internet disruption, and might see further challenges from AI. More recently, various forms of “sharing” (transportation and lodging, for example) also have emerged, and might see further changes from AI. 

But manufacturing and pharmaceuticals seem poised for AI disruption as well. On the other hand, construction might see relatively low amounts of disruption. 

Industry	Potential AI Impact	Drivers
Manufacturing	High	Robots can handle repetitive tasks, improve precision, and optimize production processes. AI can also be used for predictive maintenance and quality control.
Transportation	High	Self-driving vehicles, logistics optimization, and automated traffic management are all powered by AI.
Customer Service	High	Chatbots and virtual assistants can handle routine inquiries, freeing human agents for complex issues.
Finance	High	Algorithmic trading, fraud detection, and risk assessment can be significantly enhanced with AI.
Healthcare	High	AI can assist in medical diagnosis, drug development, and personalized medicine.
Retail	Medium	AI can personalize recommendations, optimize inventory management, and automate tasks like pricing and promotions. However, the human element in customer service and product selection might remain crucial.
Legal	Medium	AI can analyze legal documents, predict case outcomes, and streamline research tasks, but human judgment will likely remain essential for legal proceedings.
Education	Medium	AI-powered tutors can personalize learning experiences, but human teachers will likely remain central for guidance and social interaction.
Media & Entertainment	Medium	AI can personalize content recommendations and automate content creation tasks.
Construction	Low	Manual labor and on-site decision-making are still crucial aspects of construction, making widespread AI adoption less likely. However, AI can be used for design optimization and project management.
Hospitality	Low	The human touch remains essential in hospitality, but AI can automate tasks like booking and guest communication.
Arts and Culture	Low	Human creativity and emotional connection are central to the arts, making AI unlikely to replace artists entirely. However, AI can be used for artistic exploration and content creation tools.

Right now, attention is logically focused on industries and functions that are susceptible to AI automation. But equally big changes could come if AI allows competitors to enter markets in new ways. Think ridesharing and peer-to-peer lodging. 

And there always is the possibility that new industries are born. Think search and social media.

"Doom Loops" and Legacy Product Declines

It is not always easy to explain why some ideas and terms emerge at specific times in history. But some terms, including the “doom loop,” have emerged before. A doom loop is a self-reinforcing cycle of negative events that can lead to a catastrophic outcome. 

For long-time observers of the cable TV business, perhaps that phrase is current because a major cable operator now believes “the video product is no longer a key driver of financial performance.” 

source: Charter Communications 

That is a profound change for an industry known as “cable TV.” 

The “doom loop” is fundamentally caused by what Charter Communications considers an unsustainable video model.

We are familiar with the notion of the “vicious cycle,”  a situation in which one bad event leads to another bad event, which then leads to even more bad events. Then there is the phrase “death spiral,” referring to  a situation in which a company or organization is caught in a negative feedback loop, itself another phrase expressing a similar idea. 

In the environmental area, the 18th century Malthusian trap argued that population growth will eventually outstrip the carrying capacity of the environment, leading to widespread poverty and starvation. 

In recent decades we have heard about “the tragedy of the commons,” where individuals acting in their own self-interest deplete a shared resource. 

These days, we are apt to hear the term applied to the declining linear video subscription business. But we also have seen similar ideas expressed form time to time about specific companies in the telecom or connectivity business. 

source: NextTV, MoffatNathanson

One example of a doom loop is the Greek debt crisis. In 2010, Greece's government debt was too high and the country was unable to pay its debts. This led to a loss of confidence in the Greek economy, which caused the value of the Greek currency to plummet. This, in turn, made it even more difficult for Greece to pay its debts, and the cycle continued.

Another example of a doom loop is the 2008 financial crisis. The crisis began when the housing market in the United States collapsed. This led to a loss of confidence in the financial system, which caused banks to become more cautious about lending money. This, in turn, made it more difficult for businesses to get loans, which led to a decline in economic activity.

In the telecom industry, a doom loop can occur when a company's financial problems lead to service cuts, which in turn lead to customer losses, which further worsen the company's financial problems. This can create a vicious cycle that is difficult to break.

One example of a doom loop in the telecom industry is the case of Sprint. In the early 2000s, Sprint was one of the leading wireless carriers in the United States. However, the company began to struggle financially. 

In an attempt to save money, Sprint began to cut back on its network investment and service offerings. This led to customer losses, which further worsened the company's financial problems. Sprint eventually filed for bankruptcy in 2012.

Other leading firms also have experienced doom loops. MCI once was a leading provider of long distance services in the U.S. market, second only to AT&T. But MCI began to suffer as its shrinking long distance business was not offset by growth in local access services. MCI eventually was absorbed by Worldcom, which itself collapsed. 

AT&T faced the same problem, more than once. Its declining long distance business could not be countered by new revenues in local services. Eventually, after spinning off mobile, equipment manufacturing and Bell Labs assets, AT&T was acquired by SBC, which then rebranded itself as AT&T. 

Of course, a doom loop is not necessarily fatal for the company or industry in the loop. 

AT&T has a history of getting caught in doom loops. In the early 2000s, for example, the company acquired several large cable companies in an attempt to become a one-stop shop for telecommunications services and solving its local access business problem. 

However, the acquisitions were expensive and led to a significant increase in AT&T's debt. As a result, the company was forced to cut costs and lay off employees, which further damaged its reputation. In 2005, AT&T spun off its cable business.

Then AT&T decided to reimagine itself along the lines of Comcast, and acquired DirecTV and Time Warner assets. That required taking on so much debt that eventually AT&T had to sell off those assets to pay down debt. 

It perhaps goes without saying that such terms as “doom loop” only arise in connection with legacy businesses that are declining. 

By definition, growing businesses are in positive feedback loops; virtuous cycles or experiencing scale or network benefits. So you will not hear anyone applying such terms as “doom loops” to artificial intelligence.

NextLight Grabs 60-Percent Market Share Competing Against Lumen and Comcast

NextLight, the electrical utility owned internet service provider in Longmont, Colo. says it has gotten 60 percent take rates for its fiber-to-home service, with similar take rates among business customers, after gaining about 54 percent take rates after five years of operation. 

Should many other competitive ISPs achieve such success, incumbent telco and cable operator ISPs could face serious challenges. 

It has been conventional wisdom in U.S. fixed network markets that two competitors are a sustainable market structure, typically featuring one cable operator and the legacy telco, with market shares ranging between a 70-30 pattern (where the telco only has copper access)  to something closer to 60-40 as a rule (where the telco is upgrading to fiber access). 

Telcos hope for market shares approaching 50-50 as FTTH becomes the dominant access platform over time. 

The new issue is additional providers, ranging from municipal or utility-owned ISPs to independent ISPs, including independent ISP operations that cover only parts of a metro area. In a sense, that is the mass market or consumer version of the competitive local exchange carrier strategy adopted decades ago, where suppliers target business customers in major office parks or downtown core areas. 

The American Association of Public Broadband cites 750 municipal internet service provider networks in operation in the United States, mostly serving smaller communities. Not all have full retail operations, though. 

Chattanooga Electric Power claims 175,000 customers in the Chattanooga, Tennessee area. The next-largest 10 such ISPs have fewer customers, often because they are smaller population centers. 

• City of Salem Electric Department (Oregon): 50,000

• City of Longmont Power & Communications (Colorado): 40,000

• Plum Creek Electric Cooperative (North Dakota): 35,000

• Jackson Energy Authority (Tennessee): 25,000

• City of Holyoke Municipal Light Department (Massachusetts): 20,000

• City of Boulder Municipal Electric Utility (Colorado): 18,000

• City of Dubuque Utilities (Iowa): 17,000

• City of Lawrence Public Utilities (Kansas): 15,000

• City of Lexington Utilities (Kentucky): 15,000

 

And other networks are launching in larger population centers. As with any set of contestants in any other industry, not all suppliers will succeed and not all will likely survive. Managerial skill still seems to matter, as do the other prosaic concerns such as managing debt burdens and picking the right areas to serve. 

Many for-profit ISPs now believe they have better opportunities in rural areas, for example, where a new fiber network can be “first” to serve the market. Up to this point few have attempted to compete in a major big city market. ISPs targeting operations in mid-size cities have generally only chosen to serve portions of their cities. 

The obvious broader issues are the roles and strategies traditional retail service providers can envision as their markets are reshaped by competition, new investors and virtualized or other roles beyond the traditional vertically-integrated model. 

The question naturally arises: how many of these new competitors will succeed, and what are the implications for sustainable market shares over time?

In a market with two significant suppliers, each serving the whole market, an ISP might require  market share of at least 30 percent to be sustainable. That has often been the pattern where a cable operator competes against a telco with copper-only access, where the available telco speeds are quite limited in comparison to a cable operator hybrid fiber coax network. In such cases, there is an order or magnitude or two orders of magnitude difference in top speeds. 

In a market with three significant suppliers, an ISP typically needs to have a market share of at least 20 percent to be sustainable, if competition across the full geography is envisioned. Such ISPs also tend to require more efficient operations. 

In a market with four significant suppliers, where we can assume as many as two of the four compete only in a portion of the metro market, an ISP typically needs to have a market share of at least 10 percent (of the full area potential market) to be sustainable, though ISPs serving only a portion of a metro area also probably need take rates higher than 10 percent in the areas they do choose to serve. 

If an independent ISP cannot get 20 percent to 30 percent take rates in its chosen geographical areas of coverage, it probably is not doing well. 

The best suppliers can take so much share from the incumbents (telco and cable) that severe damage to the incumbent business model is possible, turning those competitive areas into loss-making operations. 

A fixed network operator with sufficiently offsetting performance might survive actual losses in a few geographies. In fact, traditional monopoly fixed network suppliers expected permanent losses in rural areas, breakeven or slightly better performance in suburbs and most of the profits from operations in city cores. 

NextLight seemingly has avoided issues of cross-subsidization of internet access service by the electrical utility ratepayers, separating its financial operations from those of Longmont Power Company.

NextLight has its own board of directors, management team, and accounting system.

NextLight seemingly provides service “at cost,” plus a small margin to cover its operating expenses. The objective is to break even, rather than “making a profit.”

NextLight's network is physically separate from LPC's network, though critics might argue NextLight uses power company rights of way and other benefits of having a sponsor with an on-going business, which could translate to financial advantages. 

Others might argue there is some cross subsidy. There is a no-recourse surcharge on LPC's electric bills, used to fund the construction and operation of NextLight, and it is applied to all LPC customers, regardless of whether they subscribe to NextLight service.

That said, NextLight has gotten a legal opinion from the Colorado Attorney General's Office stating that NextLight is not engaging in cross-subsidization, and that the non-bypassable surcharge is a fair and reasonable way to fund the network. 

In fairness, what revenue-generating entity would not look to leverage its current assets to create new lines of business? Cable operators used their video subscription networks to create fully-functional telecom networks; use their fixed network to support their mobile service provider operations; extended their consumer networks to provide business-specific services; used their linear video customer base to leverage a move into content ownership. 

Telcos do the same, when trying to extend their core operations to new services. In the more-regulated era, they had to establish separate subsidiaries to enter non-regulated lines of business. That is less an issue in today’s largely-deregulated markets. 

The city of about 100,000 is about 30 miles north of Denver, so might be considered a suburb by some, a neighboring city by others. Using either characterization, population density varies quite substantially. 

The population density of Longmont, Colorado in its city core is 11,999 people per square mile while the population density of the outlying areas is 1,369 people per square mile  

 

Housing density and population density obviously are key indicators of potential access network cost and revenue possibility. Housing density enables and constrains home broadband market size, while population density is correlated with business revenue potential. 

To a large extent, housing and population density also affect network cost: the lowest-cost-per-passing networks can be built in dense areas while the most costly networks are in rural areas. 

Among U.S. internet service providers, the “average housing density is 400 locations per square mile, with Comcast sitting squarely on that level of density. Smaller telcos tend to serve more-rural areas and have housing densities an order of magnitude or two orders of magnitude less than the largest ISPs. 

Company	Housing Units	Average Housing Density (dwellings per square mile)
Verizon	58.2 million	1,500
AT&T	51.8 million	1,300
Lumen (formerly CenturyLink)	25.7 million	600
Charter	22.9 million	500
Comcast	19.5 million	400
Windstream	14.8 million	300
Brightspeed	1.9 million	40

At least historically, that explains why Verizon was early to invest in fiber to home facilities. It has the most-dense serving areas, so has the best economics. Only recently have many smaller and independent ISPs been able to make a business case for investing in FTTH in rural and exurban areas, though lots of small rural telcos have been doing so for years. 

Housing density	Cost per home passed
40 homes per square mile	$2,000
40 homes per square mile	$800
1,300 homes per square mile	$500

Figures of merit for FTTH construction might range from $1,000 to $1,250 per household at 400 homes per square mile but $1,500 to $2,000 per household at 40 homes per square mile, for example. 

At higher densities of 1,300 homes per square mile, costs might range from $500 to $750 per household. 

The business case also includes less revenue per account potential at lower densities as well. 

All that matters as attacking ISPs and infrastructure investors weigh their odds of success when competing with legacy service providers. To be sure,  FTTH payback models seem to have changed greatly since 2000. 

The economics of connectivity provider fiber to the home have always been daunting, but they are, in some ways, more daunting in 2022 than they were a decade ago. The biggest new hurdle is that expected revenue per account metrics have been cut in half or two thirds. That would be daunting for any supplier in any industry. 

These days, the expected revenue contribution from a home broadband account hovers around $50 per month to $70 per month. Some providers might add linear video, voice or text messaging components to a lesser degree. 

But that is a huge change from revenue expectations in the 1990 to 2015 period, when $150 per customer was the possible revenue target.  

You might well question the payback model for new fiber-to-home networks which assume recurring revenue between $50 and $70 per account, per month, with little voice revenue and close to zero video revenue; take rates in the 40-percent range; and network capital investment between $800 and $1000 per passing and connection costs of perhaps $300 per customer. 

In the face of difficult average revenue per account metrics, co-investment and ancillary revenue contributions have become key. Additional subsidies for home broadband also will reduce FTTH deployment costs. 

The point is that FTTH revenue models, and the ability to sustain a competitive ISP operation, either as an incumbent or attacker, now seem to make possible more competition than was previously thought possible. 

NextLight is a good example.