The U.S. is in the middle of a digital infrastructure revolution. Artificial intelligence, cloud computing, and edge technologies are transforming industries and redefining whats possible, from national security to personalized medicine. But as AI headlines focus on coders and cutting-edge tech, the real story is unfolding in workshops and job sites where skilled workers are making innovation physically possible.
Unlike the dot-com boom or the mobile era, this AI-driven transformation isnt just about servers and software. Its about the concrete, steel, cables, power, and cooling systems that serve as the nervous system of our digital society. As the demand for hyperscale data centers and energy-intensive computing capacity grows, so does our dependence on a rising class of tradespeople who are building this infrastructure from the ground up. The future of AI doesnt just sit in a data center. Its built by hands that wire, weld, and maintain it.
WHATS DRIVING DEMAND
AI isnt a theoretical frontier anymore. It’s here, its scaling, and its accelerating the need for purpose-built facilities that can handle the load. The rise in generative AI and machine learning workloads has triggered unprecedented demand for data center capacity across the U.S. According to a 2024 report from McKinsey & Company, U.S. data center power demand is expected to more than triple by 2030from 25 gigawatts in 2024 to more than 80 gigawatts[DA1] [KG2] underscoring the urgent need to expand our physical infrastructure and the skilled workforce behind it.
This infrastructure doesnt build itself. Every new data hall or edge facility depends on a coordinated force of electricians, welders, fiber installers, HVAC technicians, and other tradespeople who bring these environments to lifeon time and to spec. Tripling power demand in just six years isnt just a tech challengeits a labor and infrastructure mandate.
THE RISE OF THE NEW-COLLAR WORKFORCE
This growing sector of workers is part of the new-collar workforcea class of skilled professionals who blend technical know-how with practical, hands-on experience. These are not white-collar or blue-collar jobs. They’re something new.
New-collar jobs typically dont require a four-year degree but demand rigorous training, problem solving, and adaptability. Theyre high-impact roles that are essential to Americas competitiveness in the AI ageand they come with real staying power. According to the World Economic Forum Future of Jobs Report 2025, frontline jobs including construction are among the fastest growing in the world, and are expected to remain in demand through 2030 and beyond.
These are careers, not just jobs. And yet, we face a looming labor shortage crisis.
SKILLED, ESSENTIAL, AND IN SHORT SUPPLY
Americas talent pipeline for the skilled trades is under severe strain. Many of the professionals powering todays infrastructure boom are nearing retirement and too few young people are being trained to take their place. Outdated perceptions about vocational training and a college-or-bust mindset have led to chronic underinvestment in trade education.
We urgently need to rethink how we train, attract, and retain critical frontline workers. That means renewing support for vocational schools and community colleges, modernizing apprenticeship programs, and changing how we talk about trade careers in the digital age. It also means building partnerships between industry and educators that deliver real-world pathways to meaningful employment.
But talk is cheap. We have built the demandnow we need the workforce to match it. Thats why my company, Compass, and our industry partners are working with Texas State Technical College to establish the MEI Data Center Program, a replicable model for nationwide workforce development. It is a hands-on, curriculum-driven initiative that equips students with the real-world skills required to launch careers in the data center industry.
WHY THIS MATTERS NOW
The future of AI, edge computing, and cloud innovation depends not only on breakthroughs in silicon or software, but on the physical infrastructure that makes those breakthroughs usable at scale.
Data centers are no longer just tech assetstheyre critical infrastructure. Just like our power grids, water systems, and communication networks, they must be resilient, redundant, and ready to support mission-critical workloads. And without a robust, future-ready labor force to build and maintain them, innovation will stall.
This isnt just an economic challenge, its a national security issue. The global race for AI dominance will be won not only in R&D labs but on construction sites and in control rooms across the country.
A NATIONAL CALL TO ACTION
We cant solve this problem in silos. Building the engine of the AI age requires a coordinated, nationwide commitment. Government, industry, and education leaders must come together to invest in the new-collar workforcebefore the gap becomes a chasm.
That means funding technical education. It means telling a different, better story about skilled trades as pathways to success. And it means recognizing and honoring the people who make our digital lives possible. We cant automate our way out of the skilled labor shortage. We need to attract, train and invest in the people who will literally build our future.
The AI revolution may be powered by machines, but its built by people. Its time we started acting like it.
Chris Crosby is the founder and CEO of Compass Datacenters.
Today, design drives effective business strategy, but design education hasnt caught up. As companies scramble to digitally transform, adapt to the climate crisis, and navigate culture and trade wars, designs role has expandedmoving to the center of how organizations shape products, services, and systems. With this elevated role comes a sobering reality: Many design leaders feel increasingly out of their depth.
Promoted for creative excellence, they suddenly find themselves navigating boardrooms, budgets, business models, and organizational change without the proper preparation. As Fast Company puts it, a generation of design leaders are in the midst of a big design freak-out, as many realize the creative confidence that propelled careers doesnt always translate into executive credibility. One senior design leader recently admitted on LinkedIn that they were unprepared to lead people, lead change, transform processes, make sound business decisions, or even understand how a company works.
This isnt a failure of individual designers. Its a failure of the system that educates them.
WHERE TRADITIONAL DESIGN EDUCATION STOPS SHORT
Traditional design programs excel at teaching craft: visual communication, UX research, design thinking. But they rarely prepare graduates for the realities of organizational leadership; topics like business strategy, change management, or stakeholder alignment are often considered outside the domain of design education.
Its not a matter of neglect. Its a matter of scope. Undergraduate- and graduate-level design programs arent meant to produce executives, just as undergraduate business degrees dont turn students into CEOs. Those leadership capabilities are built over time, and often require further training later in ones career. Yet while business leaders have long had access to MBAs, corporate academies, and executive development programs, design has had no equivalent, until now.
THE RISE OF EXECUTIVE DESIGN EDUCATION
Recent years have seen a wave of new programs spurred by this education gap. Indeed, my own employer, iF Design, last month launched the iF DESIGN ACADEMY. Drawing on decades of global design authority, the Academy develops leaders who combine creative excellence with business fluency. Our courses push participants to build skills in leadership, strategy, sustainability, and emerging technology. The goal is simple: Help mid- to senior-level design leaders grow into the role todays landscape requires.
FOUNDATIONS OF EFFECTIVE LEADERSHIP
These programs arent about turning designers into MBAs. They aim to cultivate a hybrid mindsetone that blends creativity with executive acumen. Core skills taught include:
Understanding how businesses create and measure value
Communicating with influence across organizational functions
Navigating metrics, org structures, and operational complexity
Driving change in environments that resist it
Leading teams with psychological safety and purpose
These arent nice to have skills, but the foundations of effective leadership in any domain. Doug Powell, lead lecturer at the iF DESIGN ACADEMY and former VP of design at IBM, captures the challenge well: While many design leadership courses focus on the management and performance of the teamwhich is critically importantmy course focuses on the skills, behaviors, and tactics of navigating the broader ecosystem of leadership in complex organizations. This outward focus is too often dismissed in leadership training, but without these essential skills, leaders will continue to struggle.
DESIGNS POWER IS REALBUT ONLY WITH LEADERS WHO CAN HARNESS IT
McKinseys widely cited 2018 report, The Business Value of Design, found that companies in the top quartile of design performance outpaced industry benchmarks by as much as 2:1. Good design drives growth, customer loyalty, and competitive advantage, but only when leaders embed it into business strategy. Design thinkers often serve as translators, connecting human-centered thinking to business outcomes. To succeed in this task, design leaders need more than intuition, they need executive fluency. As Katrina Alcorn, managing director at Accenture Song, put it: The backlash against design is a de facto backlash against innovation. As Alcorn reminds us in her piece Good Design Is (Still) Good Business, companies that slash design roles for short-term savings risk long-term irrelevance. Good design creates real business value, but only if design leaders have the authority and tools to lead.
Today, design extends far beyond aesthetics. It shapes how organizations think, operate, and deliver value in a volatile world. That responsibility grows larger every year, and our investment in design leaders must grow with it. Executive design education is not a luxuryit is essential to unlocking designs full power to drive progress, resilience, and a better future for all.
Lisa Gralnek is global head of sustainability and impact for iF Design, managing director of iF Design USA Inc., and creator/host of the award-winning podcast, FUTURE OF XYZ.
According to the National Association of Corporate Directors, boardrooms today face a dizzying list of risks: economic volatility, geopolitical tensions, cybersecurity threats, technological disruption, and a tightening labor market. But the one risk too often overlooked? That businesses rely on healthy people and healthy communities.
Despite spending more on healthcare than any other nation, the U.S. is falling behind on nearly every major health indicator. Life expectancy is declining, chronic illness is rising, and access to care remains uncertain for one in four Americans. These arent just public health issues. Theyre economic issues. They weaken our workforce, strain businesses, threaten national security, and erode trust in institutions.
The equation is simple: Healthy communities fuel healthy businesses. One Deloitte report estimates that improving health across the U.S. could add $2.8 trillion to GDP by 2040, with corporate profits possibly increasing by $763 billion. In todays environment, companies prioritizing health attract better talent, earn more trust, and stay more competitive. Its a business risk no leader can afford to ignore.
FROM CHARITY TO STRATEGY
For too long, corporate social responsibility was treated as an afterthought. A check written here, a charitable initiative there. But stakeholders are demanding more. A recent survey found that 84% of Americans believe corporations have a responsibility to strengthen the communities where they operate, and 72% say those corporations should help solve major systemic issues.
Today, employees, customers, and investors expect it. Businesses are being judged not only by quarterly earnings, but by how they show up in communities. In a polarized world, trust is fragile. And once lost, it is hard to regain. Thats why business leaders can no longer view community health as charity. It must be seen for what it is: strategy.
WHAT WORKS
At CHC: Creating Healthier Communities, Ive seen firsthand what happens when companies get it right. Weve had the privilege of working with businesses that are moving beyond charitable donations to co-lead real solutions: Weve partnered with Ameriprise Bank to host a series of workshops to advance mental wellness in the workplace. Ardelyx is facilitating community engagement activities to increase access to services. The Samaritan Health Project, Inc. hosted health fairs and connected residents to pharmacies that provided discounted rates on prescriptions. Hilti launched Mental Health Mondays for employees. These leaders recognize the truth: When community health declines, so does the bottom line.
These efforts succeed because they are local, collaborative, and sustained. They arent acts of charity. They are smart investments in a healthier, more productive future.
COLLABORATE FOR GREATER IMPACT
In the last century, value was measured almost exclusively in financial terms, such as quarterly returns, market share, and shareholder wealth. But that equation is shifting. Today, the true currency of competitive advantage lies in the ability to collaborate across boundaries, earn trust in a skeptical world, and harness data for collective impact.
Thats the vision behind our new Leadership Council for Healthier Communities (LCHC)the first national council of its kind designed to bring leaders from business, philanthropy, health systems, and grassroots organizations together to cocreate solutions and measure results.
LCHC isnt about replacing what companies are already doing. Its about connecting, aligning, and scaling those effortswhether thats addressing maternal health, tackling obesity and cardiometabolic disease, strengthening nutrition and food security, or ethically leveraging new tools like artificial intelligence to improve access to care.
In short: Its a place where organizations across sectors can collaborate to turn commitments into outcomes, and strategy into results.
A STRONGER FUTURE, TOGETHER
Declining community health isnt an abstract concernits already hitting the bottom line. And unless businesses act, the costs will only grow.
But the reverse is also true. When companies invest in healthier communities, employees thrive, talent pipelines expand, and customer trust deepens.
The companies that thrive in the next decade will be those that treat community health as strategynot as philanthropy or PR.
That work cant happen in silos. It requires leaders willing to collaborate across sectors, share what works, and hold themselves accountable for results. The next generation of value creators will be those who partner across boundaries, invest in the health of people and places, and make trust their competitive edge.
Thats the vision behind the Leadership Council for Healthier Communities, a platform where business leaders can scale what works and unlock growth that benefits everyone.
When we invest in healthier communities, we dont just create stronger neighborhoodswe create stronger businesses, stronger economies, and a stronger future. Because the health of business and the health of communities are inextricably linked.
Jean Accius, PhD is president and CEO at CHC: Creating Healthier Communities.
I recently spoke to a donor who reviewed a batch of proposals from different groupsdifferent names, different logos, but nearly the same projects. Teams had reinvented the same wheel in parallel. Individually, some of those projects might get funded. Collectively, the sector missed the chance to pool efforts and solve a larger piece of the problem. That felt wrong, not because anyone was bad, but because our systems make it easier to duplicate than to unite.
Heres what should terrify donors: Even as funding tightens, duplicated projects still get financed while collaborative funds report backing organizations that figured out how to work together, with $23 billion in annual support.
WHY WE BUILD IN SILOS
First, complexity encourages focus. Its human to take the problem you can manage and try to perfect it. Building a consortium feels messy and risky. Second, funding mechanics favor tidy, single-org proposalsgrant windows, scoring rubrics, and reporting templates push teams into solo asks. Third, survival instincts kick in: Leaders protect jobs and short-term stability. Fourth, donors often prefer neat comparables; its easier to evaluate and communicate about a single-organization proposal than to underwrite a messy, multipartner bet.
Those incentives produce dozens of similar pilots, fragmented data, duplicated engineering, and wasted momentum. The cost isnt just administrative friction; its slower adoption of useful tools and weaker influence at policy tables where a unified voice would matter.
EVIDENCE: JOINT BETS WIN
There were dozens of AI-for-good panels and side sessions at the United Nations General Assembly this year, but no single calendar. Everyone scrambled, duplicating efforts, and wasting hours trying to track what was happening where. We published an open list of eventsa basic piece of shared infrastructureand the relief was immediate. If we cant coordinate a calendar, how will we synchronize shared models, data standards, or joint deployments across countries?
This isnt abstract. Two other practical cases show what coordinated work unlocks.
The Systems Transformation Hub saw several climate organizations realize they were spending most of their time fundraising rather than changing policy. They pooled efforts, cocreated a shared roadmap, and unlocked far more capital and traction than any member could have achieved alone. That shift, from isolated asks to a joint strategy, created real forward movement.
Googles Flood Hub teaches a related lesson on adoption: Building a model is not the same as embedding it into systems. Even with brand and engineering muscle, tech needs coordinated deployment. Governments, local NGOs, and scientists formed a collaborative around the tool so it could be used operationallynot just demoed. That adoption work turned capability into impact.
WHAT REAL COLLABORATIVES MUST DO
Collaboration is not logos on a slide. It is actual working alliances built on practical choices:
Shared infrastructure, not replicated widgets. Build once, deploy many timescommon models, APIs, and clear data contracts so local teams adapt rather than rebuild.
Coordinated fundraising, not competing pitches. One unified ask with transparent allocation reduces transaction costs and unlocks deeper capital.
Backbone governance. Someone must hold procurement, compliance, and opsthe boring glue that keeps partners moving.
Shared measurement and safe data practices. Established metrics and data agreements let partners iterate and credibly report joint outcomes.
Funders who reward collectives. Change scoring rubrics to favor system change, and long-term stewardship over tidy outputs.
These are governance decisions. Theyre not glamorous, but they determine whether pilots plateau or scale.
WEVE DONE THIS WRONG, TOO
Im not writing from a pedestal. At Tech To The Rescue, we have made the same mistakes, matchmaking that at times created more iterations than systems. We also see the alternative: projects that plug into shared services, adopt common standards, and scale faster. The tech side is also unlocking quickly. Many off-the-shelf tools now cover a large share of operational needs, but that potential only becomes impact if the sector coordinates adoption. Thats why were convening partners this season to test an AI-for-good ecosystem, a minimum viable product for shared infrastructure, pooled fundraising, and collective accountability. Its operational testing, not optics.
DONORS: CHANGE THE INCENTIVES
Donors, this piece is for you. If you fund this space, please examine whether your forms, timelines, and scoring favor tidy, single-org asks or whether they actively reward collective proposals, backbone support, and multi-year, flexible capital. Consider creating grant lines specifically for collaboratives. Pool funds with other donors to underwrite backbone organizations; fund the integration workprocurement, ops, governancethat rarely looks glamorous but makes scale possible. Also consider directing funds toward policy, watchdogs, and capacity-buildingthe systems layer that keeps shared tech safe, accountable and effective.
If your organization wants impact more than visibility, ask whether the next grant helps the sector or mainly your institution. If you build tech, ask whether your next sprint will be useful to 10 deployments, and if so, design it to be shared.
We have the tools, talent, and urgency. The remaining barriers are cultural and structural: the humility to share credit, governance to coordinate, and funding models to pay for integration. Because the problems we care about wont wait while each of us finishes our own version of the same thing.
Jacek Siadkowski is cofounder and CEO of Tech To The Rescue.
On the heels of Starbucks’ recent announcement it will be cutting 900 corporate roles and closing 1% of its Northern American stores by the end of 2025 (after accounting for both new openings and closures), Starbucks Workers United said Tuesday that 59 of those locations marked for closure are unionized locations.
Starbucks Workers United, the worker-led effort to unionize Starbucks baristas, represents 12,000 baristas in 45 states and Washington D.C., across more than 650 cafes.
The closures, announced last week by CEO Brian Niccol, are part of a massive $1 billion restructuring strategy dubbed Back to Starbucks, aimed at turning around declining sales and brand image damage.
Representatives from Starbucks and Starbucks Workers United declined to share details about the fate of specific locations.
While we remain outraged at how callously Starbucks handled these closures, we are proud that we have forced the company to make this process fairer for impacted union baristas, said Starbucks Workers United spokesperson Michelle Eisen. These measures to support baristas show the power and strength of our union. We remain focused on organizing stores and demanding Starbucks settle a fair union contract that improves hours and staffing, increases take-home pay, and resolves unfair labor practices.”
Reached for comment, a Starbucks spokesperson told Fast Company: “Given the industry-leading offer provided to all affected partners including reassignment opportunities where possible as well as generous severance we were able to quickly reach an agreement with Workers United to similarly help represented partners through this transition. This reflects our commitment to partner care.”
A spokesperson for the company also told amNY that unionization was not a factor in the decision to close specific locations.
Baristas from closing stores will either be offered severance packages or transferred to new locations, which has led uneasy employees to crowdsource their own list of shuttering locations as they wait for official word.
The stakes remain high for Starbucks if it fails to settle a contract and workers go on strike ahead of the holiday season, which is the busiest and most profitable time of the year.
U.S. job openings were essentially unchanged million last month amid economic uncertainty arising from President Donald Trumps trade policies and an impending government shutdown.
The Labor Department reported Tuesday that job openings blipped up to 7.23 million from 7.21 million in July. Economists had forecast a drop to 7.1 million.
The Job Openings and Labor Turnover Survey (JOLTS) showed that layoffs fell month. But so did the number of people quitting their jobs which is a sign of confidence in their prospects of finding a better job. The report’s measure of hiring last month was the weakest since June 2024.
Job openings remain at healthy levels but have fallen steadily since peaking at a record 12.1 million in March 2022 as the U.S. economy roared back from COVID-19 lockdowns.
The U.S. job market has lost momentum this year, partly because of the lingering effects of 11 interest rate hikes by the inflation fighters at the Federal Reserve in 2022 and 2023 and partly because Trumps trade wars have created uncertainty that is paralyzing managers trying to make hiring decisions.
Altogether, Tuesday’s JOLTS numbers suggest that the job market remains in an awkward place: Americans who have jobs are mostly safe from layoffs. Unemployment remains low at 4.3%. But jobseekers are struggling to find work.
Companies are clearly hoarding workers with the economy still at full employment, Carl Weinberg, chief economist at High Frequency Economics, wrote in a commentary. “It will take a bigger blow than what we have seen so far to convince companies that it is safe and prudent and necessary to lay off workers.”
Labor Department revisions earlier this month showed that the economy created 911,000 fewer jobs than originally reported in the year that ended in March. That meant that employers added an average of fewer than 71,000 new jobs a month over that period, not the 147,000 first reported. Since March, job creation has slowed even more to an average 53,000 a month.
On Friday, the Labor Department is expected release numbers on September hiring and unemployment though the report could be postponed if a budget impasse in Congress leads to a government shutdown Wednesday.
If the report comes out, it is expected to show that employers added 50,000 jobs in September, unimpressive but up from a meager 22,000 in August, according to a survey of economists by the data firm FactSet.
At their last meeting two weeks ago, Fed policymakers cut their benchmark interest rates for the first time this year to support the sputtering job market. They also signaled that expect two more rate cuts this year.
Paul Wiseman, AP economics writer
If it seems like it’s getting more expensive to replace a broken door, kitchen fixtures, or upgrade a major appliance, youre not wrong.
The cost of home repair and remodeling projects is up compared to a year ago and running ahead of inflation overall, according to a report from data analytics company Verisk.
The firm’s latest Repair and Remodeling Index jumped 3.4% in the April-June quarter compared to the same period last year. That’s a bigger annual increase than the 2.7% rise in inflation in the same period, as measured by the Consumer Price index.
The index, which tracks costs for more than 10,000 home improvement products, including appliances, doors, plumbing, and windows, showed a roughly 0.6% increase from the January-March quarter.
While costs did continue to rise, they rose at a slower rate than in the first quarter,” said Greg Pyne, vice president of Pricing at Verisk Property Estimating Solutions.
Much of the increase in home repair and remodeling costs appears to be driven primarily by higher labor costs for repair and remodeling work, Verisk noted.
The second-quarter jump in costs for home improvement products coincided with the Trump administrations broad rollout of tariffs on imported goods from many of the nations major trading partners. But the tariffs didn’t have the expected impact given they were postponed several times and didn’t fully take effect until early August, midway through the third quarter.
However, homeowners looking to replace cabinetry could soon see prices increase sharply, following a new volley of tariffs announced by President Donald Trump last week that includes a 50% import tax on kitchen cabinets and bathroom vanities due to kick in on Wednesday. John Lovallo, an analyst with UBS, estimates the tariffs on cabinets and vanities could add roughly $280 to the cost of a home.
The most labor-intensive types of home repair or remodeling work registered the biggest quarterly increases in labor costs. For example, the cost of replacing tile flooring rose 1.2%, while the cost of remodeling a primary bath or replacing vinyl siding each rose 1% in the April-June period from the previous quarter.
Nearly all of the 31 categories of repair and remodeling work tracked by Verisk saw costs increase at least slightly.
The latest index puts costs for repair and remodeling at almost 62% higher than they were 10 years ago and more than 73% higher than the first quarter of 2013, when the index debuted.
After declining the past two years, homeowner spending on maintenance and home improvement projects increased in the first half of this year, according to researchers at Harvard University.
The universitys Joint Center for Housing Studies most recent leading indicator of remodeling activity, or LIRA, estimates spending hit $510 billion in the second quarter, a 1.8% increase from a year earlier. However, the researchers project that growth in spending on home improvement and maintenance will slow in 2026, citing weakness in the housing market and slower construction of new homes.
The housing market has been in a slump since 2022, when mortgage rates began climbing from historic lows. Sales of previously occupied U.S. homes sank last year to their lowest level in nearly 30 years. And, so far this year, sales are running below where they were at this time in 2024.
Alex Veiga, AP business writer
Curing cancer. Reducing carbon emissions. Maximizing business efficiency. To achieve all this and develop untold social goods, artificial intelligence accelerationsts at companies like Google, Meta, and OpenAI believe their industry has a duty to speed ahead towards superintelligence, or AI thats far superior to humans at most tasks. Key to that revolution will be the build-out of data centers.
Meanwhile, a technical transformation of the workplace already appears to be underway. The nations largest employer, Walmart, said that because of its AI implementation, hiring will remain flat over the next three years even if revenues rise. Every businessnot just the big oneswill eventually reckon with some version of that transformation. Whoever wields the technology best will get an edge. Regulators, in turn, must find forward-looking ways of controlling the excesses of the winners while mitigating the hardship of the losersand fast.
Sen. Mark Kelly fears that the biggest losers could be working-class people. The Arizona Democrats AI for America plan, arguably the most comprehensive Democratic answer to the Trump administrations pro-industry AI Action Plan, would create an industry-financed AI Horizon Fund to pay for energy-grid upgrades and workforce reskilling.
But while Kellys plan is admirable …
it dodges the policy specifics necessary for real legislation. He also fails to grasp certain economic and political realities of the AI industry and its players. And the federal government, as it heads for a shutdown, seems far from capable of passing any thoughtful AI legislation.
Here, we attempt to fill in these gaps.
Data centers everywhere
The AI models poised to reshape business practices reside on servers humming away in the dark within massive single-story buildings called data centers.
New data centers represent the most tangible sign of the so-called AI boom. Most estimates say that there are more than 500 hyperscale data centers, housing tens or hundreds of thousands of servers, in operation in the U.S. today. Between 50 and 100 more are either licensed or under construction in 2025.
Kellys home state of Arizona is regarded as one of the most attractive places for data center projects because of the low cost, reliable power, affordable land, easy permitting, and tax incentives. (Apple, Google, Microsoft, and Amazon Web Services (AWS) already have data centers there.)
States compete to attract data center projects. They come at a cost. Over the past five years, 10 states have already lost more than $100 million per year to data center tax abatements, with Texas and Virginia each giving away roughly $1 billion, according to a study by Good Jobs First, an economic development policy advocacy group.
According to the study, a total of 32 states offer such exemptions to Big Tech companies and their partners; 12 states dont disclose the exemption amounts, which makes calculating a national total difficult. But its in the billions, and climbing. Whether all this investment truly delivers in the long run remains unclear.
The infrastructure gap
All this is causing Arizona citizens to ask more about these projects. In August, Tucson rejected Project Blue, a proposal to build a 290-acre AWS data center near the city. They cited concerns over water use, the potential burden on the local power grid, and the possibility of spiking electricity rates to fund additional power infrastructure. Deloitte estimates that the power demand from AI data centers in the U.S. could grow to about 123 gigawatts by 2035, up from roughly 4 gigawatts required in 2024.
The problem is that the existing power grid was built to serve households and businesses, not legions of sprawling data centers. When a new one goes up, the local or regional energy supplier typically must augment the capacity of the grid to meet the demand. And those infrastructure costs are often passed on to residents through rate hikes or tax increases. Those same tax increases and electricity rate hikes could hit businesses in the area, including small businesses.
Who should pay?
Kelly believes that AI companies should pay for energy infrastructure upgrades necessitated by data center power demand. But his proposal offers no mechanism for metering the AI companies financial obligation or amount they should pay into the fund for infrastructure augmentation.
Making this workable would require working with utilities and state and local energy regulators to determine a fair fee. To pay for infrastructure upgrades, Kelly could require a small but significant percentage of every megawatt of power purchased by the data center operator to go into a hypothetical fund.
Congress could also require data center developers to buy or lease enough land to contain both their facilities and the renewable energy infrastructure to power and cool them. The data center operators could also be required to pay to connect the renewable sources to the local grid, should the power they generate go unutilized. Elon Musks xAI, for example, brought its own power to its massive Colossus data center in Memphis. Unfortunately, it was dirty powermethane-powered turbines, and the facility quickly became one of the areas biggest pollutersa cautionary tale.
For a city and its utility, the biggest fear is that an AI data center could pick up and leave, in pursuit of more permissive environmental laws or cheaper power rates, leaving behind an empty hulk and suddenly unemployed local workers. Establishing federal-level environmental guidelines and power-grid responsibilities could remove some of the incentives to leave, forcing data-center operators to consider that at least some of those costs would be the same no matter where they went.
Reskilling, but make it AI
Tech companies often say that in their ideal world, humans will work in tandem with AI tools, and that new jobs will emerge that require some skill with these technologies as old ones are eliminated. Arriving at the right balance will likely take years. Because of the ongoing, rapid advances in AI, the process may never truly end.
In the near term, the biggest beneficiaries are likely to be the companies selling the tools. Kelly argues, reasonably, that the AI companies should help pay for the costs of job displacement and reskilling workers. He suggests that the AI Horizon Fund be used to pay for AI education programs at community colleges, trade schools, and universities.
Kelly also believes that the government should pass laws to make sure that workers themselves benefit from AI efficiencies. This could mean reimagining what the workweek looks like, as well as policies to strengthen worker bargaining power through stronger union representation.
In the early days of the internet, collectors traded rare whiskey and wine on eBay alongside Beanie Babies and vintage sneakers. But then, in 1999, six months after closing down firearm sales, eBay announced they would ban the sale of alcohol and tobacco products as well.
“As a general rule, these laws are just so complex and contradictory, that we just decided that in the best interest of our users to prevent that situation from ever occurring,” then-spokesman Kevin Pursglove said.
More than 25 years later and almost a century after the end of Prohibition, the regulatory environment is no less forgiving, and the resale of spirits online has been scattered across niche forums, gray-market Facebook groups, and high-end houses like Christies, Sothebys and Bonhams.
The patchwork of U.S. liquor laws
Domestic laws are complex. Six states still ban retail spirits sales on Sundays, to start. Seventeen operate as control states with a government monopoly on liquor sales. And while 47 states allow wineries to ship directly to consumers, only 11 extend that same privilege to distillers, according to the Distilled Spirits Council of the United States. Industry groups have shepherded a bill into Congress that would give the USPS authority to mail alcoholic beverages, but a final vote could take years.
Few competitors have cracked this market. BAXUS uses blockchain for tokenized bottle trading, while U.K.-based Whisky.auction, Whisky Auctioneer, and Whisky Hammer focus outside the U.S. Good Bottle Auctions, based in Connecticut, sticks to the Northeast with in-person delivery.
But since launching in 2020, a company called Unicorn has expanded to serve the continental U.S., logging over six million bids across half a million lots, totaling $125 million in sales. Its Chicago vault holds more than $100 million in inventory, with weekly and monthly auctions that ship purchases to pickup locations nationwide.
Sothebys vs. Unicorn
CEO Phil Mikhaylov, who previously worked at UberEats and delivery startup GoPuff, frames Unicorns speed and scale as a direct extension of that background. He says that most legacy auction houses top out at a few hundred lots per week, while Unicorn regularly clears thousands.
I think what you see at other auction housessay, a Sothebysis they might have roughly 300 lots in an auction every single week. We’re doing, on a bad week, 3000 lots. On average, we’re doing 4000 to 5000 a week, Mikhaylov says.
Unicorn CEO Phil Mikhaylov [Photo: Unicorn]
Sothebys, for the record, did $114 million in spirits and wine sales in 2024a slump from their record of $159 million the year before.
At this point, we sell $12 million of whiskey per week, Mikhaylov said in an interview earlier this year with their newly acquired (and renamed) in-house magazine, the Unicorn Review.
Whats happened is that weve become effectively a redistribution platform. A bottle thats selling for $50 in Chicago might have $200 of value to someone in New Mexico.
Unicorn has spent yearsand millions of dollarsbuilding the framework to do so. As of 2025, they have 12 dropoff points across the country, which then link out to a handful of vans that drive across the region, all eventually relaying bottles back to their Chicago processing facility. Sellers can hand off their bottles, and buyers can drive out to any of the facilities, or go through Unicorn for delivery.
For legal reasons, Unicorn (a licensed retailer) must take possession of every bottle and bring it to its central Chicago location. There, an authentication and appraisal team runs each one through what may be the most robust spirits-resale database in the industry, producing estimates that cover everything from a 1982 Chateau Lafite Rothschild Bordeaux to a jug of Jose Cuervo.
I tested the system by submitting four bottles for appraisal online. The sheet broke down batch numbers, year, and proof, and offered estimates within a few dollars of their eventual sale price. After handing the bottles off the next week in-person, they were on auction by Sunday: one sold $5 below estimate, another $5 above, and the other two squarely within range. Mikhaylov credits the accuracy to their in-house tracking software and the size of their data set.
A data-driven bar cart
Digitizing that many bottles was slow at first, but after five years (and some AI integration), the system has basically been perfected. Bottles are scanned, authenticated, photographed in 360 degrees, and logged into the vault. Customers using Unicorn for storage pay 25 cents per bottle per month and can check their inventory online at any time.
[Photo: Unicorn]
Storage complements sales: Mikhaylov says that many collectors were looking for a solution to manage their treasure troves, often resorting to home-brewed spreadsheets and lists. Now, whales might have their entire collection picked up, palletized, digitized, and tracked onlineready for observation, admiration, or sale at auction.
If they do decide to sell, payouts are delivered in 10 days or less, minus $5 per bottle and 5% commission (buyers pay a 15% premium). Compared to traditional auction houses, which often require weeks or even over a month to process consignments and payout sellers, thats fast.
The idea began during the pandemic. Mikhaylovs cofounder, Cody Modeer, found himself unable to auction bottles from his shuttered bar; traditional houses dismised his inventory as too low-value. Meanwhile, Mikhaylov was searching for a way to digitize and manage spirits collections, allowing collectors to drop off bottles anywhere in the country and track them remotely.
There wasnt a modern platform that was meeting the demands of today’s consumer: something that made it easy to drop off, to sell, or to manage a collection, Mikhaylov says.
This is meeting the demands of that consumer. 70% of our clientele that signed up this year have been Gen Z and millennial. For the younger demographic, its all about transparency and speed . . . Think of us like Kelly Blue Book, but for your bar cart.
U.S. stocks are coasting toward the finish of Wall Streets latest winning month on Tuesday.
The S&P 500 fell 0.2% in afternoon trading but remains on track for a fifth straight winning month after setting a record last week. The Dow Jones Industrial Average was down 145 points, or 0.3%, as of 1:43 p.m. Eastern time, and the Nasdaq composite was 0.3% lower.
Oil-related companies weighed on the market after the price of crude fell again as traders see too much oil washing around the world. Schlumberger fell 3.8%, and Halliburton dropped 3%.
They helped offset a 12.7% jump for CoreWeave, which said Meta Platforms will pay up to $14.2 billion for a new order for cloud computing power made under its existing service agreement, with the potential for more.
Treasury yields eased in the bond market following a couple mixed reports on the U.S. economy. One said consumers are feeling less confident than economists expected, with many respondents in the Conference Board’s survey pointing to the slowing job market and inflation that has remained higher than anyone would like.
A second report suggested the job market may be remaining in its low-hire, low-fire state. U.S. employers were advertising roughly the same number of job openings at the end of August as the month before. The hope on Wall Street had been for a number that’s neither too high nor too low, one balanced enough to keep the Federal Reserve on track to continue cutting interest rates.
The Fed just delivered its first cut of the year, and officials have penciled in more through the end of next year to give the job market a boost. If data on jobs come in too strong, it could make the Fed less willing to cut rates. If the numbers are too weak, meanwhile, they could mean a recession is coming.
Either extreme would hurt the stock market, which has run to records from a low in April in large part on expectations that the Fed will cut rates several times. The stock market is already facing heavy criticism for being too expensive after prices ran so high.
Another potential wild card is hanging over the market, meanwhile. The U.S. government seems to be heading toward a shutdown at the end of the day following another political impasse in Washington.
The economy and stock market have made it through past shutdowns without much wear, and many economists and professional investors feel relatively OK about another one. The S&P 500 has climbed an average of 4.4% during past shutdowns and is positive over the last five, according to Monica Guerra, head of U.S. policy at Morgan Stanley Wealth Management.
The timing of this potential shutdown, though, would likely cause delays for several important economic reports. That includes a release due on Friday about how many jobs U.S. employers created and destroyed in September.
That could make Wall Street twitchier when investors are already nervous about the state of the economy and what that means for the potential for cuts to rates. The Department of Labor has already said that the Bureau of Labor Statistics will completely cease operations if theres a lapse.
On Wall Street, Spotify Technology sank 6.4% after the Stockholm-based streaming giant said its founder, Daniel Ek, is stepping down as CEO to become the executive chairman. Two of his lieutenants will replace him as co-CEOs: Chief Product and Technology Officer Gustav Söderström and Chief Business Officer Alex Norström.
Lamb Weston jumped 4.1% after the supplier of frozen French fries and other potato products reported a stronger profit for the latest quarter than analysts expected.
In stock markets abroad, indexes ticked higher in Europe following a mixed finish in Asia.
In the bond market, the yield on the 10-year Treasury eased to 4.14% from 4.15% late Monday.
Stan Choe, AP business writer
AP Business Writers Yuri Kageyama and Matt Ott contributed.
