AI’s Power Demand Is Driving Up Your Electricity Bill

Tech companies are racing to build new data centers to handle the surge in AI computing demand. These facilities use massive amounts of electricity — just this year, electricity demand from data centers in the U.S. is estimated to rise 22%.

The problem is that the electrical grid isn’t built for that kind of growth. Power shortages are already delaying data center projects by two to six years. Which raises two questions: Can the U.S. actually produce and deliver enough electricity to sustain this boom — and if it can, who’s paying for it?

Nowhere is that question more extreme than with OpenAI. OpenAI wants to build out a chip network that would consume 250 gigawatts (GW) of energy by 2033. That is equivalent to a fifth of America’s entire electric generation capacity.

The U.S. added a record 56 GW of new power capacity last year. Even if that pace continued annually, OpenAI’s plan would require the company alone to use more than half of all the new electricity the country adds over the next eight years.

Where would that power come from? Nuclear is the popular talking point, but to generate 250 GW from nuclear alone, you’d need roughly 250 new reactors at an estimated cost of $12.5 trillion. Each one can take 15 years or more to build, so the math and the timeline don’t add up.

Natural gas isn’t a realistic option either, as much as Trump wants it to be. Just 12.5 GW of natural gas capacity is expected to come online this year, one of the biggest additions ever. OpenAI’s target is 20x that.

The obvious answer is solar. Renewable power like solar and onshore wind is the least expensive and quickest power generation source to deploy in the U.S., even without government subsidies. Of the 56 GW of capacity added last year, more than half of it came from solar.

But politics are getting in the way. The One Big Beautiful Bill Act scraps key tax credits and adds new permitting hurdles, including a requirement that new projects be personally approved by Interior Secretary Doug Burgum. These changes could cut U.S. solar capacity projections by roughly 20%.

Energy isn’t the only challenge. Local communities are increasingly pushing back against new data centers. About $64 billion worth of projects have been blocked by residents who don’t want them in their communities. Their criticism is simple: These projects raise costs without giving much back.

In major data center hubs, electricity prices have jumped as much as 267% over the past five years, according to Bloomberg. And despite the huge price tags, they don’t create many jobs.

For many communities, the math just doesn’t work: higher electricity bills, environmental impacts, and little economic benefit — all to power the infrastructure behind an AI boom that may be running faster than the grid, the economy, or the public can keep up with.

If I had to predict what will happen here, it’s that there will be a huge investment in the grid, the majority of which will be absorbed by AI. 

It’ll be positioned under the auspices of innovation and future and maintaining our lead, but the net effect will be taxpayers, most of whom are not benefiting from the run-up in AI stocks, will end up subsidizing these companies’ actual energy costs.

The fact that energy prices are going up highlights how AI is actually expensive. It actually costs something, and someone has to pay for it. And up to this point, it has seemed to most people that it’s just these like, you know, Qatari billionaires or these Silicon Valley billionaires are just funding these things. A trillion dollars here, a trillion dollars there. 

But we’re really not in a place of abundance in the way that Sam Altman seems to think. We have a scarcity of energy to the point that the costs are going to be shouldered by regular Americans.

Roughly 1,500 new data centers have been announced this year. The operative word here is announced. 

Of all of the data centers that have been announced — only a quarter are under construction. When you look at the earnings of the companies that actually build the data centers, the skyrocketing demand ... just isn’t there. 

Fluor Corp., one of the top data center construction companies, just reported a 6% decline in its sales and a 12% decline in its backlog (money tied to contracts that have been signed but not yet fulfilled). And as Herb Greenberg noted, demand for Mueller Industries’ copper tubing, a key component in data center HVAC systems, has been muted — well below what you’d expect in the middle of a supposed data center boom.

This could be an expected demand problem or an energy problem, but what it certainly is is a threat to the greater economy. AI is the single-largest source of investment debt in the U.S. If investors start questioning future returns and ask for higher yields, the whole AI bubble could burst.

The conversation around technology and the role that it’s having on young people’s mental health has taken center stage. According to the CDC, adolescent depression is up nearly 60% since 2010, tightly correlated with smartphone adoption, and Meta’s own research found that nearly one-third of teen girls say Instagram worsens their body image. 

In response, tech companies have rolled out new “safeguards.” Meta recently introduced new parental controls on Instagram, and in September, Sam Altman announced a 120-day plan to roll out similar safeguards for ChatGPT. That pledge followed a lawsuit from parents who accused OpenAI of contributing to their teen’s suicide.

But the follow-through on these announcements has been disappointing. Last week, Altman said that OpenAI has “been able to mitigate the serious mental health issues” such that it can relax some of the previous restrictions. Meta’s track record isn’t better — Research shows that its “parental safety” features have been largely ineffective. 

Meanwhile, children are becoming more exposed to digital risks. Teens spend five hours/day on social media — 35% of their waking hours outside school. This pattern looks like it will continue as 29% of U.S. kids ages 8 to 10 years old own a smartphone.

Kids are a hard audience to lose. Facebook, Instagram, Snapchat, TikTok, X, and YouTube earned nearly $11 billion from U.S. users under 18 in 2022. To put that in perspective, $11 billion could: 

So the question stands: Are these companies making meaningful changes for teen mental health, or just protecting a very profitable status quo?

We nod our heads while leaders like Sam Altman speak in hushed tones about being concerned with the commonwealth, when in reality their first, second, and third priority is shareholder value. The thing is, that’s his job. It’s our job to elect leaders who say plainly that kids should not be in these relationships and should not be consuming content that has been shown to drive a 60% increase in self-harm.

When I look at the mental health debate in tech, what stands out is that these companies have a lot to say about it, but their actions show they are not serious. 

There are effective ways to deal with this: Norway has banned social media for kids under 13 and plans to raise the age to 15. Australia just passed a law banning social media for children under 16. 

That is the kind of clarity that works. What does not work is waiting for CEOs to wake up one morning and decide their priority is child well-being over shareholder value, because that will never happen.

____________sponsored content ____________

Create standout content quick

Now, everyone in marketing, HR, and sales can create their own on-brand content. With Adobe Express, it’s quick and easy to create campaigns, videos, social posts, presentations, pitch decks, and more. Brand kits and locked templates make following design guidelines a no-brainer. And commercially safe AI features, powered by Firefly, let teams create with confidence. So your brand shows up consistently — everywhere.

Learn more at: https://www.adobe.com/express/spotlight/business?sdid=6S3T6V56&mv=display&mv2=display

Sponsored by Adobe Express

____________sponsored content ____________

Back in July, we called SpaceX the most important monopoly no one’s talking about. It effectively owns the modern space economy. 

SpaceX launched 1,060 spacecraft in the second quarter of this year, accounting for 88.5% of all launches worldwide, and Starlink represents roughly 65% of all active satellites in orbit.

But that monopoly may finally be under threat. A new generation of challengers is emerging, each trying to chip away at SpaceX’s lead in launch, connectivity, and orbital infrastructure.  Investors have recognized the potential here: AST SpaceMobile is up 230% in the past year, and SpaceX launch rival Rocket Lab is up 500%.

Meet the competitors:

AST SpaceMobile is trading at 500x expected 2025 revenue ($60 million), and its stock has gone up 2,500% over the last two years. Rocket Lab is up over 500% over the past year.  

The market isn’t acting rationally with these companies, and when you look at a performance like that, you should be wary. AST SpaceMobile and Rocket Lab are just the names that are associated with space that are public, and people just want to get some sort of exposure to the industry. But this exposure is not tied to the fundamentals. To me, it’s more about momentum.

There’s one company that you’d want to invest in: SpaceX, and you can’t because it’s private, so you go to these competitors.

Friend: someone who listens, responds, and supports you. Or, a $129, always-on pendant with a microphone, Bluetooth, and a connection to Google Gemini to produce text-based responses that might sound like support. 

Unlike human friends, this Friend comes with fine print, including “biometric data consent” clauses that let the company passively capture audio and video, collect facial and voice data, and use it all to train its AI systems.

The device, created by 22-year-old Harvard dropout Avi Schiffmann, started shipping this summer. It has no screen, no camera, and no direct access to the internet. Only 1,000 units have shipped, but that didn’t stop Schiffmann from shelling out $1 million to run the biggest ad campaign in the New York City subway this year.

The ads have attracted criticism for their controversial assertion that artificial intelligence can be a friend. But setting that debate aside, Friend overlooks a key point: Humans are visual creatures. Without any kind of visual interface, it’s harder for humans to form a real connection with the device.

Our visual cortex is 3x larger than the auditory cortex and 2.5x larger than touch. Our brains process images 60,000x faster than text, and roughly 90% of all information we absorb is visual. Albert Mehrabian, a researcher of body language, postulated that 55% of communication impact comes from visual body language while intonation accounts for 38%, and words 7%. The phenomenon is so consistent it has a name: the visual dominance effect, where visual stimuli override other senses when competing for attention.

We’ve seen this preference play out over decades. The radio flourished until TV came along. Fast-forward a century, and the podcast industry, traditionally audio-only, is increasingly a visual medium. Video podcast viewership increased by 40% year over year in 2025, whereas growth in traditional audio podcast listenership was closer to 7% over the same period. That’s why Spotify recently announced a partnership with Netflix to distribute podcasts on the streaming platform.

The iPhone demolished its smaller-screen competitors, and Apple outsold the entire Swiss watch industry by putting a screen on our wrists. 

The premise of a voice-first, screenless device directly contradicts how the brain wants to receive and process information. 

We have seen this play out before. The Humane AI Pin flopped so hard that it triggered more returns than sales, and Rabbit’s viral orange AI device sold over 130,000 units, but according to the company’s own data, it has just ~5,000 daily active users — a less than 4% retention rate.

If the Friend AI necklace makes it into the tech history books, it will be for a fundamental misunderstanding of human connection. Instead of a human companion, Friend is an always-listening human spy. And not a good one at that. “The life of spies is to know,” wrote English poet George Herbert — but Friend proves how little it does by forgetting that connection starts with what we see.

Scott and Ed sat down with antitrust expert Jonathan Kanter to unpack the urgent case for regulating AI-driven tech monopolies. They dig into how dominant companies wield AI to lock in power, whether existing antitrust frameworks are up to the task, and what a truly competitive tech future might look like.