Why 99.999 Percent Availability is Not Possible Anymore

Our user experience of applications, devices and networks is far from the “five nines” standards (99.999 percent availability) telcos used to tout. 

As a practical matter, today’s heterogenous, edge-powered, internet transport fabric, IP-based application environment absolutely means user experience cannot approach 99.999-percent availability for any applications. 

That might not apply to core systems in banking, financial trading or some security-critical use cases, but only to the core systems, not the end user access of those systems. 

The problem is that no matter what any single participant in the value chain might claim for its own availability, and even if that availability is between 99 percent and 99.99 percent, the entire end-to-end value chain depends on the sum total of availability across the whole value chain, and that math is challenging. 

Consider an example where contributor availabilities are:

• Device: 99%

• Home broadband access: 99.5%

• Internet backbone: 99.99%

• App server: 99.9%

• Local power: 99.5%

The end-to-end availability requires multiplying all those discrete availabilities. So the formula is 

0.99 × 0.995 × 0.9999 × 0.999 × 0.995 ≈ 97.4 percent. That means 229 hours of downtime per year, not the 5.26 minutes per year allowed by "five nines” standard.

The only reason end users seem unaware of the change is that much of the downtime happens when they are not actively using their connections (devices not present; devices in “do not disturb” mode; user is sleeping; apps not in immediate use). 

Value Chain Component	Typical Availability (%)	Major Downtime Factors
User Devices – Mobile	95%–99%	Battery loss, OS/software issues, dropped connections
User Devices – Fixed	96%–99.5%	Power outages, device crashes, local network (Wi-Fi) issues
Access Network – Mobile	97%–99.9%	Tower outages, congestion, interference, maintenance
Access Network – Fixed	98%–99.9%	Fiber/cable cuts, power issues, last-mile failures
Global Internet Backbone	99.99%+	Rare fiber cuts, DDoS attacks, routing errors
Application Servers (Cloud)	99.5%–99.99%	Cloud region outages, software bugs, maintenance, cyberattacks
Local Power Supply	99.0%–99.9% (urban)	Grid instability, storms, infrastructure failures
End-to-End Availability	Often < 95%–98%	Cumulative failures across components

Some Problems Just Cannot be Fixed: Lumen Technologies, for Example

Some problems are nearly impossible to fix. Consider Lumen Technologies, a mashup of the former Level 3 Communications capacity business and the former US West local telco business. Right now, Lumen is facing declining revenue growth and a substantial debt load, acquired at the same time the firm made acquisitions of capacity assets growing its share of that business. 

The current business strategy seems clear enough: refocus on what used to be the Level 3 Communications business (capacity and enterprise) and eventually get out of the former local telco business. 

To be sure, the former US West (then Qwest, then CenturyLink) always was structurally challenged. The former US West had low customer density and limited business customer potential. 

The former means its network costs on a per-location business were going to be high, while the latter means its organic growth potential is limited. All the other Regional Bell Operating Companies formed during the breakup of the AT&T system had denser network footprints and higher business customer potential. 

On top of that, US West never developed a facilities-based mobility business, as did the other firms that eventually became AT&T and Verizon. And since mobility now is the revenue driver for a local telco, that also has hampered Lumen’s growth. 

Today, the enterprise business represents about 75 percent of total Lumen revenues.

The issue is how Lumen could divest essentially all of its former local telco business, and to whom. Private equity is a possibility. Perhaps some rural telcos or independent broadband internet providers could buy parts of Lumen. 

Perhaps some assets could be sold to new public-private partnerships, cooperatives or other joint ventures. That would be especially true of the rural assets. 

By some estimates, the metro areas of Denver, Seattle, Phoenix, Salt Lake City, Portland, Minneapolis-St. Paul, and Orlando generate between 60 percent and 70 percent of all the local telco revenues Lumen generates. 

The point is that nobody has been able to overcome the density and customer upside issues US West, Qwest, CenturyLink, Lumen have faced since the beginning. Prior to 2000, US West tried to create a position in cable TV services. Under Qwest, from 2000 to 2010 or so, the company shifted to the capacity business. 

CenturyLink seemingly wanted both scale benefits and entry into a higher business customer profile. Lumen is reversing that focus, selling off essentially all the former CenturyLink rural telco assets and now its mass markets fiber to home business. 

Lumen will become Level 3 Communications. The legacy telco assets will eventually be divested, somehow, to various buyers, in some way. It seems unlikely the whole former US West business will be appealing to any single buyer; and perhaps no single buyer will have the capital and business plan requiring all the assets, in any case. 

Some problems seemingly cannot be fixed.

Water Footprint Matters for the "Great American Desert"

It is easy to forget that the U.S. Intermountain West; the states containing 40 million people who use the Colorado River; and the Great Plains states (the land east of the Sierra Nevada and west of the 96th meridian) are essentially deserts.

And that matters because water is life. 

By now most of us are familiar with the concept that every physical object and every intangible product has both a carbon footprint and a water footprint. Most people probably do not pay much attention to water footprint. Many of us in the U.S. intermountain west and Great Plains likely do pay attention, as befits people living in deserts. 

Product	Unit	Water Footprint (approx.)
Beef	1 kg	15,415 liters 415
Pork	1 kg	5,988 liters 1
Chicken	1 kg	4,325 liters 1
Cheese	1 kg	3,178 liters 1
Eggs	1 egg	52 gallons (197 liters) 6
Milk	1 liter	1,021 liters 1
Rice	1 kg	2,497 liters 1
Bread (wheat)	1 kg	1,608 liters 1
Tomatoes (fresh)	1 kg	214 liters 7,1
Apples	1 kg	822 liters 1
Almonds	1 kg	16,194 liters 1
Chocolate	1 kg	17,196 liters 1
Coffee	1 cup	34 gallons (129 liters) 6
Wine	1 glass	34 gallons (129 liters) 6
Jeans (cotton)	1 pair	2,108–2,866 gallons (8,000–10,850 liters) 8,6
T-shirt (cotton)	1 shirt	659 gallons (2,720 liters) 8,6
Smartphone	1 device	3,190 gallons (12,760 liters) 8,6
Car	1 car	13,737–21,926 gallons (52,000–83,000 liters) 8,6
Leather shoes	1 pair	2,113–3,626 gallons (8,000–13,730 liters) 8,6
Paper (A4 sheet)	1 piece	1.3 gallons (5.1 liters) 8

The point is that in many parts of the world, water footprint matters as much as carbon footprint. 

Computing Always Has Caused Job Losses: Why Would AI be Different?

Since all prior instances of computing applied to work have eliminated some number of jobs, it only makes sense to predict that artificial intelligence also will do so. But among the questions is the issue of incidence: where the job losses will occur.

Historically, the introduction of computers and automation led to the reduction or elimination of many routine, repetitive, or data-driven job functions. And though we might be tempted to think the jobs were “blue collar,” most of the losses affected “white collar” workers in offices. 

AI seems destined to become a substitute for a wider range of “routine tasks” but also is ;going to affect more complex cognitive work as well. 

Job Title/Function	Cognitive Tasks	Risk from AI	AI Substitution
Data Analyst/Market Researcher	Data analysis, report generation, forecasting	AI excels at pattern recognition, data processing, and generating insights from large datasets	Automated data analytics, predictive modeling 1,2,3
Paralegal/Legal Assistant	Document review, legal research, drafting	AI can rapidly review documents, extract information, and draft legal texts	AI-powered document review, legal research bots 1,2,3
Content Writer/Copywriter	Article writing, editing, content creation	Generative AI can produce high-quality written content at scale	Automated content generation, editing tools 1,2,4
Entry-Level Accountant/Financial Analyst	Bookkeeping, financial reporting, auditing	AI automates repetitive calculations, reconciliations, and report generation	Automated bookkeeping, financial analysis 1,4,2
Customer Service Representative	Responding to inquiries, troubleshooting	AI chatbots and virtual assistants handle routine queries and support tasks	AI chatbots, automated helpdesks 2,4,3
Junior Software Developer	Writing/debugging code, code review	AI can generate, test, and debug code with increasing accuracy	AI code generation, bug detection 2,3
HR/Administrative Assistant	Scheduling, resume screening, workflow management	AI automates scheduling, candidate screening, and routine admin tasks	Automated scheduling, resume parsing 2,4
Middle Manager (Routine Tasks)	Report consolidation, performance tracking	AI can aggregate data, generate reports, and monitor KPIs	Automated  reporting, dashboarding 5,4
Translator	Language translation, localization	AI models can translate text and speech across languages with high accuracy	Machine translation, localization tools 6,7
Insurance Underwriter	Risk assessment, policy review	AI can analyze risk factors and automate policy decisions	Automated risk analysis, policy generation 4

Routine, structured cognitive tasks are most vulnerable, especially those involving data processing, document review, and standardized content creation. We might already guess that entry-level and support roles in law, finance, marketing, and administration face the highest risk.

Jobs requiring human judgment,  creativity or complex interpersonal skills are less exposed, but even these may see significant transformation as AI tools become more sophisticated.

The history of applied computerization also set that pattern.

Job Function	Technology	Impact/Change	Era	Source(s)
Typis, /Secretary	Word Processing, PCs	Drastic reduction; executives self-type documents	1980s–1990s	1,3
Data Entry Clerk	Databases, Automation	Largely automated; fewer manual data entry roles	1980s–2000s	2,6
Bookkeeper, Payroll Clerk	Accounting Software	Routine tasks automated; fewer clerical jobs	1980s–2000s	4,5,1,3
Switchboard Operator	Digital Telephony	Obsolete; replaced by automated systems	1970s–1980s	2,3
Travel Agent	Online Booking Platforms	Reduced demand; self-service travel arrangements	1990s–2000s	4,6
Paralega, /Legal Assistant	AI Document Review	AI handles research, document review	2020s–	6,10
Financial Analyst, Trader	AI, Algorithmic Trading	Automated analysis and trading	2010s–	7,8,6
Customer Service Rep	Chatbots, AI Assistants	AI handles routine inquiries	2010s–	1,11,2
Content Writer, Editor	Generative AI	AI drafts articles, reduces junior writing roles	2020s–	7,6,10
Junior Software Developer	AI Code Generation	AI writes/debugs code, reducing entry-level roles	2020s–	9,10