Dan J. Harkey

Master Educator | Business & Finance Consultant | Mentor

The Speed of Change Is Accelerating

Are we ready, and willing participants or laggards?

by Dan J. Harkey

Share This Article

Summary

Computing’s exponential trajectory (often framed through Moore’s Law) has underpinned decades of disruption.  That change now collides with an AI build‑out whose energy and land requirements are reshaping real estate, labor markets, and electric grids.  Global data‑center electricity demand could more than double by 2030, with the U.S. accounting for a large share of the growth; grid access is increasingly the limiting factor.

Meanwhile, many work activities are technically automatable, creating real displacement risks if national leadership fails to act.  My article defines the phenomenon, quantifies the pressures, and lays out solutions that eschew participant politics in favor of pragmatic national strategy.

1) Defining Moore’s Law

Moore’s Law is the empirical observation that the number of transistors on an integrated circuit has historically doubled roughly every two years, a trend first articulated by Intel cofounder Gordon Moore in 1965 (initially one year, later revised to two).  It is not a Law of physics, but a planning heuristic that captured decades of manufacturing and design innovation across semiconductors.

The Law’s essence—compounding improvements in cost, density, and performance—helped drive down “quality‑adjusted” prices for compute, memory, and sensors and enabled successive waves of digital transformation.

2) Technology Has Disrupted the Status Quo in Manufacturing and Services

Over the past decades, automation and digitization have steadily reallocated tasks from humans to machines, both in manufacturing (robots, CNC, MES) and in services (software, RPA, analytics).  Contemporary analyses find that today’s technologies could, in theory, automate more than half of U.S. work hours, with the most significant exposure in routine cognitive tasks—admin, some legal drafting, document preparation, and parts of finance and sales—while hands-on care and high‑judgment roles remain less exposed.  That is a measure of the potential for task automation, not a deterministic job‑loss forecast, but it signals the breadth of disruption already underway.

Historical evidence shows displaced workers often suffer persistent earnings losses absent targeted retraining and placement, underscoring why traditional “status quo” policies have struggled to cushion shocks from technology and trade.

3) Now, AI Wants a Significant Slice of Jobs

Recent research frames the future of work as a Partnership between people, software “agents,” and robots.  Within that frame, roles at the highly automatable end may comprise ~40% of total jobs, concentrated in administrative, some legal, and specific machine‑operation activities.  Again, these are technical potentials; the realized outcomes hinge on adoption speed, workflow redesign, and policy choices.

AI’s Impact on roles and the creation of new jobs underscores the importance of leadership actively supporting workforce transitions through reskilling and job creation, inspiring a sense of shared responsibility.  

4) AI Data Centers: Hyperscalers Consuming Land and Electricity Like No Other Enterprise

The AI wave is driving a hyperscale land‑and‑power race.  North American vacancy fell to record lows (≈1.6–3%), absorption hit multi-gigawatt levels, and construction backlogs doubled amid power constraints and interconnection delays.  Site selection increasingly pivots on available grid capacity and large contiguous parcels.

Electricity demand is the central bottleneck:

  • The IEA projects global data‑center electricity consumption rising from ~415 TWh (2024) to ~945 TWh by 2030—more than doubling—led by the U.S. and China.
  • DOE’s 2024 assessment estimates U.S. data centers consumed ~4.4% of national electricity in 2023 and could reach ~6.7–12% by 2028, reflecting AI, cloud, and electrification trends.
  • Market reports show preleasing >80% and record absorption in Northern Virginia, Atlanta, Chicago, and Phoenix, with newcomers (West Texas, Louisiana, Indiana) rising due to land, tax incentives, and power accessibility.

At the hardware layer, AI accelerators (e.g., NVIDIA H100) can draw ~700 W per GPU; millions deployed translate to multi-gigawatt continuous loads, highlighting why rack densities and cooling (increasingly liquid) are transforming facility design.

The AI hyperscale data centers’ power demands are projected to reach multi-gigawatt levels by [2030], emphasizing the urgent need for policymakers to address grid capacity and future energy security.

5) “Move Over, Workers—Your Job Has Gone”: A Provocative Claim, but Outcomes Are Not Preordained

The U.S. has a super high demand for skilled trades in the foreseeable future. Please read my article about the high demand.

https://danharkey.com/post/skilled-trades-surge-americas-quiet-workforce-revolution 

It is technically feasible to automate large shares of routine work, and some roles will shrink.  However, rigorous reviews caution against assuming a one-way path to mass unemployment; outcomes depend on skill transitions, workflow redesign, and demand growth in complementary activities.  Evidence from displaced-worker programs also warns that generic retraining is often insufficient unless it is targeted, credentialed, and aligned with employer demand.

Policy implication: The choice is not binary (jobs vanish vs. status quo).  Instead, nations must actively manage transitions through targeted retraining, credentialing, and employer-aligned programs to minimize economic scarring and maximize productivity gains.

6) “For Those Who Are Left, Plan on Paying Double for Your Electricity.”

Electric bills are rising and expected to outpace inflation through 2026, driven by demand growth (including data centers), regional grid upgrades, and fuel dynamics.  Regions like the Pacific and New England face sharper increases, emphasizing the need for region-specific energy policies and infrastructure investments.

That said, localized effects can be sharper where AI load growth collides with tight capacity, and some utilities report significant capital needs that can pressure rates.  Technical reports document CapEx for distribution up ~50% (2019–2023) and note state-level spikes tied to new industrial and data‑center loads.

The AI demand for land and electricity Appears Daunting, but with targeted reforms in permitting, interconnection, and transmission, policymakers can effectively expand data centers without destabilizing the grid.

To address the looming constraints, policymakers should prioritize reforms in permitting, interconnection, and transmission, including incentives for grid upgrades and GETs, to ensure AI data centers can expand without destabilizing the grid.

Given the scale of AI-driven displacement, federal programs like WIOA must be expanded and tailored with automation-specific tools, wage insurance, and transition accounts to support displaced workers and prevent prolonged economic hardship effectively.

7) Leadership Blind Spots: Failing to Recognize “Light‑Speed” Acceleration of Change

Energy policy commentators and former FERC leadership note that the U.S. must balance AI competitiveness, affordability, and reliability, and are urging rulemaking to streamline extensive load connections while protecting ratepayers.  Without proactive, coordinated action, localized crises—such as price spikes, curtailments, and community pushback—are plausible.

8) Solutions: An Agenda for Effective National Leadership (Foregoing Participant Politics)

Below is a non-partisan blueprint that operates across supply, grid, demand, and workforce.  It is evidence-based and implementation-focused.

A) Expand Clean, Reliable Generation and Storage

·       Fast‑track dispatchable low‑carbon capacity (uprates in existing nuclear; firmed renewables with storage; flexible gas with CCS where appropriate), coupled with long‑duration storage pilots for AI‑heavy regions.

·       Behind-the-meter microgrids at data centers (combining renewables, storage, and peaking assets) to reduce grid stress and provide grid‑support services (frequency response, black‑start).

B) Unlock Transmission and Interconnection

·       Permitting reform and corridor pre-designation for high‑voltage lines; mandate utility adoption of GETs (dynamic line ratings, topology optimization, advanced power flow controllers) to squeeze 10–30% additional capacity from existing networks while new lines are built.

·       Standardize and digitize interconnection queues for large loads and onsite generation; create fast-track pathways for projects meeting efficiency and flexibility criteria. 

C) Make AI Loads Grid‑Friendly by Design

·       Obligate flexibility: require hyperscale campuses to support demand response, scheduled training windows, and geographic load‑shifting (a significant share of AI compute is timing‑flexible).  Tie the interconnection priority to quantified flexibility commitments.

·       Set efficiency baselines: national PUE (Power Usage Effectiveness) and liquid‑cooling standards for high-density AI halls; accelerate heat‑reuse to district systems.  Support the proposed Liquid Cooling for AI Act to drive R&D assessment and federal best practices.

·       Transactive energy pilots: compensate data centers for real-time grid services via market products, aligning AI operations with grid stability goals.

D) Price Signals That Protect Households

·       Rate design guardrails: ring‑fence household affordability (lifeline blocks, arrearage management) while exposing large new loads to marginal‑cost‑reflective rates that incentivize flexibility and onsite investment, consistent with EIA’s outlook on regional price pressures.

·       Targeted bill credits in high-impact zones funded by load‑flexibility revenues and public-private grid investments, preventing de facto cross‑subsidies from households to hyperscalers.

E) A National Workforce Transition Compact

·       Wage insurance + reemployment accounts: pilot earnings‑top‑up for displaced workers transitioning to lower‑pay roles, paired with time-bound reemployment accounts for accredited training and job‑placement services (building on evidence about persistent earnings losses).

·       Apprenticeships at scale: expand Registered Apprenticeship into AI‑adjacent fields (data‑center operations, power electronics, HVAC‑liquid cooling, transmission construction), leveraging WIOA funding channels and employer consortia.

·       Regional Transition Hubs: co-locate community colleges and employer labs with grid and data‑center projects; prioritize programs with verified placement rates and portable credentials rather than generic retraining.

F) Measurement, Transparency, and Governance

·       National AI‑Energy Observatory at DOE: continuous reporting of AI energy use, grid impacts, and efficiency progress; publish campus-level flexibility scores and PUE/heat‑reuse metrics to benchmark operators.

·       Scenario planning: adopt IEA’s global demand forecasts and RAND’s barrier taxonomy to stress‑test state plans; require utilities to model AI load cases in Integrated Resource Plans (IRPs).  

9) What This Means for America

  • Jobs: Without deliberate action, task automation could outpace worker transitions.  With the Compact above, the country can convert displacement risk into productivity gains and better-paid technical careers around the AI‑energy nexus.
  • Electricity Bills: National averages likely rise, but not inevitably “double.” Leadership can decouple household affordability from hyperscale demand via rate design and flexibility obligations.
  • Power & Land: AI campuses will keep expanding.  The question is whether they arrive as grid assets (flexible, efficient, co-investing) or grid burdens.  Policy design decides.

10) Closing Argument: Lead the Acceleration

Moore’s Law taught industry to plan for compounding.  AI now forces the government to do the same—for energy, infrastructure, and human capital.  The agenda above is neither partisan nor speculative; it is operational and grounded in current demand trajectories, grid realities, and labor‑market evidence.  If national leadership acts with urgency, America can win the AI race, keep power affordable, and bring workers along rather than leaving them behind.

Quick Reference (Data Points & Sources)