How Electricity Charging Might Change

It now is easy to argue that U.S. electricity pricing might have to evolve in ways similar to the change in retail pricing of communications services changed in the shift from analog to digital formats. 

Significantly, retail pricing might change from “consumption” or “usage” to “capabilities” or “access.” In other words, commercial power customers might eventually be charged based on “how much” power is available; where it is available or when it is available. 

Consider the earlier change in connectivity service pricing. 

For the most part, connectivity providers (telcos, mobile operators) no longer price their services on “usage” (minutes, calls, texts, bytes consumed), preferring “capability” and “access” as the key pricing elements. 

For internet access services, consumption does not typically matter. Instead, prices are based on “potential speed.” So a 100-Mbps connection costs the least; a 500-Mbps connection costs more while a gigabit-per-second costs the most.

Electricity still is mostly priced based on consumption (usage). But the economics of paying for the common costs of generation and transmission remain, even as more customers reduce consumption using self generation (solar panels, local generation by businesses).

Electric grid support therefore will become more challenging as user consumption drops, based on substitution of local generation for network-delivered power.

The basic business problem is that this forces a smaller number of customers to bear a larger portion of shared cost recovery, to the extent that common costs are recovered from usage charges. 

Electricity service providers have some tools to reinvent their business models. Load management becomes more important, for example. 

A shift to “access” charges also would help, creating a different model not based on actual account energy consumption but a fee based on ability to use the network. That mirrors the flat monthly fee approach now used by mobile service providers, where prices are not dictated by the number or length of phone calls, the number of text messages sent and received, or the amount of internet access data consumed. 

Instead, one fee, providing access to the network and its services, dominates. 

As with communications companies, customers who want “bigger pipes” would pay more, as do customers who want gigabit internet access service, compared to those who only want to pay for 100-Mbps speeds.

That is important in an era where local generation is going to reduce grid-delivered power consumption. 

Electricity is ceasing to be an “energy sales business” and becoming an infrastructure access business, exactly like telecom. Where “amount of electricity consumed” used to drive the revenue model, the telecom approach would substitute “ability to use the network and its features.” 

Consumer solar users without extensive battery assets then would pay for the ability to use grid power at night, for example, in the same way that a mobile device user “pays for” the ability to use the mobile operator network, rather than the specific amount of consumption of network resources. 

The alternative is continued cross-subsidy collapse, where costs keep rising for customers unable to switch to some form of self generation. 

Common costs (generation and transmission) must be recovered. Self generation threatens the present model. As with communication networks, electrical grids must be designed to support peak demand, not average demand. 

Network revenue models must assume universal service and recovery of all common costs, not simply marginal costs related to actual consumption. 

Traditional pricing assumes energy consumption is equal to grid usage. But distributed generation breaks that assumption. 

Essentially, customers remove themselves, at least partially, from the system, but retain the optionality of using the grid for reliability, backup, and peak load balancing. 

But fixed costs stay embedded in the price of per-kiloWatt hour charges, so rates will rise as sales fall. 

At the same time, new demand driven by high-performance computing and associated data centers increases the need for new investments in transmission infrastructure as well as generation, increasing the fixed costs. 

The business model will break, if not revamped.

Which Future for Neoclouds: Rational Consolidation or Collapse?

Technology market structures tend to change as they age. Small upstart companies get acquired; bigger firms merge; a few dominant leaders emerge, taking a “winner takes most” structure. 

Any market researcher, studying any particular capital-intensive market, will tend to find something like a Pareto distribution often applies: up to 80 percent of results are produced by 20 percent of actors. Some might call that the rule of three. 

Market share structures in computing, connectivity and software tend to be fairly similar: leadership by three firms, corresponding to the rule of three. 

“A stable competitive market never has more than three significant competitors, the largest of which has no more than four times the market share of the smallest,” BCG founder Bruce Henderson said in 1976.  

Codified as the rule of three, the observations explains the stable competitive market structure that develops over time, in many industries. 

Others might call this winner take all economics. 

So a logical question is what happens in the high-performance computing market, including the market space for neocloud providers such as CoreWeave, Nebius and others changing their business models from bitcoin mining to focus on artificial intelligence model training and inference operations. 

Some might argue we are shifting from a focus on training capabilities and towards inference operations. It’s hard to argue with that observation, as models become routine apps used by businesses and consumers. 

So some might argue we could see less need for highest-performance compute capabilities of the sort neocloud providers offer. Others might argue more of the computational load will be handled by edge devices, and there is some truth to that position as well. 

But inference operation ubiquity does not necessarily mean less power; less powerful chips; fewer operations inside massive data center complexes; less physical real estate or water consumption. 

Although pre-training growth is slowing, and compute is shifting from training to inference, the compute demands from post-training scaling and test-time scaling, and increased usage suggest that the world likely needs a lot more AI-focused data centers, and the ramp from US$300 billion to US$400 billion in 2025 to roughly US$1 trillion in 2028 is directionally realistic, according to one Deloitte estimate. 

So the future might not include less need for high-performance computing facilities. 

On the other hand, what technology market has not evolved over time to patterns with just a handful of market leaders?

So if the independent neocloud provider market follows the historic pattern, market consolidation will happen, with a handful of major, scaled neocloud providers; traditional hyperscalers, plus a long tail of smaller, niche players.

Some argue the process has already begun. 

But there are other possibilities as well.  The neocloud provider market might not consolidate but instead collapse.

The "Big Three" hyperscalers possess massive scale, deep financial resources, and comprehensive service portfolios, allowing them to engage in price wars and continuously innovate at a pace the smaller players cannot match. So some would argue this creates immense and unsustainable pressure on the neoclouds' margins and ability to compete effectively in the long term.

Without a genuinely-unique value proposition or niche, the independent neocloud providers might struggle to retain customers who often prefer the security and breadth of services offered by the large providers.

The hyperscalers also will be better positioned to handle the likely higher regulatory costs, ability to attract talent and risk aversion of enterprise customers, as well. 

A collapse scenario might happen for at least some providers if customers abandon the neoclouds because of longevity fears. The danger cannot be dismissed. 

That happened around the turn of the century to many would-be capacity providers and competitive local exchange carriers. 

In the late 1990s, driven by the Telecommunications Act of 1996, which opened markets to competition, hundreds of new companies rushed to build wide-area optical fiber networks and local access facilities.  

This resulted in a vast oversupply of "dark fiber" (unused capacity), with estimates suggesting 85 percent to 95 percent of constructed fiber went unused after the bust. 

The industry and investors widely believed demand for bandwidth would grow indefinitely, leading to an investment frenzy based on the mentality of "if you build it, they will come". Actual demand and revenue growth, however, did not keep pace with the rapid network construction, creating an unsustainable business model for many.

CLECs and fiber providers were able to secure massive amounts of funding through debt and speculative equity offerings. When the broader stock market began to decline in 2000, this financing dried up, immediately pushing heavily leveraged companies into bankruptcy.

Hypercompetition and Price Wars: The presence of too many competitors in the same markets led to vicious price wars that drove down bandwidth prices (in some cases, by 60 percent per year), making it difficult for many new entrants to become profitable or even cover their costs.

In that case, rational merger activity did not drive the consolidation. Instead, the sectors mostly collapsed into bankruptcy. It’s impossible to tell, today, which of these outcomes develops. Over-investment, over-capacity and inadequate demand have happened with many earlier technologies, including railroads in the nineteenth century; the telecom and internet bubbles of the late 1990s and early 2000 era.

How AI Could Affect Your Investing Strategies

If you are active as an investor, you've had to spend at least some time evaluating where and how to participate in artificial intelligence: what to buy, what to avoid, and your reasons for doing so. And some of the implications are a bit startling for our thinking about computing-related hardware and software.

Generative AI might turn some computing “principles” upside down, while sustaining others. We have in recent decades seen software produce more value than hardware. In place of asset-light software, we might see value created in greater amounts by capital-intensive physical infrastructure. 

Examples might include compute “as a service” providers; power providers; fiber networks and cooling solution providers. Returns might flow to a smaller number of suppliers able to afford the huge investments in capital-intensive, long-lived physical facilities underpinning AI compute operations. 

Asset-light software might produce less value. Contrary to the recent “software eats the world” model, AI rewards scale and capital access. 

And where value has been created by asset-light, fast-moving small teams, AI should favor larger providers with enough scale to navigate markets that are highly-regulated. 

Regulatory compliance and trust barriers will tend to protect incumbents with scale. 

Likewise, we might see a shift in acquisition value. Where merger and acquisition activity recently has been about “acquiring talent,” AI might force something of a shift to “acquiring assets.”

That might include sources of proprietary data, distribution capabilities and relationships or compute infrastructure and energy resources, rather than teams of people. So the “aqui-hire” strategy might have to be revised. 

On the other hand, generative AI might support the current value of “distribution” or direct customer relationships. Much as distribution became more important once the cost of creating content dropped (social or legacy media), so, as content creation increasingly has a marginal cost of production near zero, 

audience control captures value.

Much of the impact of computerization in general, and AI in specific, has been to emphasize value creation underpinned by scarcity, on one hand, and by scale on the other hand. This sort of “high and low” or “barbell” source of value squeezes out the middle (good but not great; too much labor to fully automate, not enough brand equity to command premium pricing). 

But, in some cases, the changes will be dramatic. Where business strategy, until recently, was to “move up the stack” from lower levels to higher, the reverse could happen, in some instances. 

Value and competitive moats might be created “down the stack” in infrastructure, rather than “up the stack” in apps. “Asset ownership” might produce more value than “asset-light” business models. 

Value also might hinge, in some cases, on better applied judgment (figuring out the better models, sources of value and sources of scarcity (data, distribution, regulatory barriers). In at least some cases, that might mean a revenue model based on outcomes or performance. 

Industry	Value Chain Role	Judgment Being Scaled	Why It Wins	Likely Monetization
Professional services (legal, accounting, consulting)	Senior advisory / opinion layer	Risk tradeoffs, precedent weighting, strategic advice	Execution automates; clients still pay for responsibility	Outcome fees, retainers, premium advisory
Healthcare	Diagnostics , treatment planning	Pattern recognition + clinical judgment	AI assists, but liability and trust anchor value	Per-decision, subscription to clinicians
Finance / Investing	Portfolio construction, risk oversight	Capital allocation under uncertainty	Alpha = judgment, not data volume	Assets under management fees, performance fees
Insurance	Underwriting, pricing	Risk selection and exclusion	Better judgment = structural margin advantage	Loss-ratio-driven profits
Cybersecurity	Threat prioritization , response	Signal vs noise discrimination	Attack volume explodes; prioritization is scarce	Platform + premium response services
Media, content	Editorial direction / curation	What matters, what to ignore	Abundance makes selection valuable	Subscriptions, sponsorships
Education	Curriculum design, assessment	What to learn, in what order, and why	Content cheap; sequencing is hard	Tuition, cohort-based pricing
Supply chain, logistics	Network design, exception handling	Tradeoffs between cost, speed, resilience	Automation fails at edge cases	Optimization-based pricing
Enterprise IT	Architecture, systems integration	Tradeoffs across cost, security, flexibility	Complexity increases with AI	Long-term contracts
Telecom / connectivity	Network planning, traffic engineering	Capacity allocation under uncertainty	AI drives demand volatility	Regulated or contract pricing
Energy, utilities	Grid management, load balancing	Reliability vs cost vs emissions	Errors are catastrophic	Regulated returns
Marketing, growth	Strategy, budget allocation	Channel mix, attribution judgment	Content automates; spend decisions don’t	Performance-based fees
E-commerce, retail	Merchandising, pricing strategy	Demand forecasting, margin tradeoffs	SKU explosion increases complexity	Margin expansion
Manufacturing	Process optimization, quality control	Yield vs throughput tradeoffs	AI reduces waste; judgment prevents failure	Cost savings share
Real estate, infrastructure	Capital allocation , siting	Location and timing decisions	Long-lived assets amplify good judgment	Asset appreciation
Regulatory, compliance	Policy interpretation, enforcement	Ambiguity resolution	Rules expand faster than clarity	Subscription + advisory