Artificial Intelligence

[phkrause]

AI perception bubble

Illustration: Allie Carl/Axios

Simon Hernandez-Arthur — the new author of Axios Communicators, our weekly newsletter covering the biggest trends in the comms industry — writes: AI has become an everyday tool for communicators, but many of the audiences they're trying to reach remain deeply skeptical of how it's being used. Why it matters: Communicators risk getting stuck in a pro-AI bubble, where the ubiquity of the technology among colleagues and executives makes it easy to overlook the skepticism — and sometimes outright hostility — it generates. Half of U.S. consumers prefer to give their business to brands that don't use generative AI in consumer-facing messages, advertising and content, according to a Gartner survey from earlier this year. Younger consumers were particularly skeptical: 57% of Gen Z and millennials said generative AI has made content quality worse. Keep reading.

[phkrause]

🤖 The new existential threat to AI

Illustration: Sarah Grillo/Axios. Stock: Getty Images

Forget energy. Forget chips. Forget China. The most clear and present danger to AI and any AI-related economic boom is rapidly rising public opposition to U.S. data centers, Jim VandeHei and Mike Allen write in a "Behind the Curtain" column. Why it matters: Republicans and AI CEOs are in full panic mode watching politicians and the public turn on the physical engines of AI growth. In private, they tell us they can't find a compelling message to shift opinion fast enough. Their nightmare — a political and economic daisy chain that cripples the AI revolution — is already taking shape: Voters revolt against data centers over electricity bills, water use, giant warehouse-like developments and fears about AI eliminating jobs. Democrats race to capitalize, competing to impose tougher restrictions or block projects altogether. Republicans — watching candidates get punished in places like Ohio — retreat from an industry they've spent years championing. The pipeline of new data centers slows, constraining the computing power AI companies need to keep improving their models and products. Investors question the staggering spending, borrowing and valuations built on expectations of relentless AI growth. 🔭 Zoom out: The AI boom has become dangerously intertwined with the U.S. economy, meaning this will ricochet far beyond Silicon Valley. Goldman Sachs estimates U.S. AI investment will hit about $600 billion this year, equal to roughly 2% of GDP and 10% of business fixed investment. Big Tech has already locked itself into trillions of dollars in future spending: The Financial Times found six hyperscalers have nearly $1.5 trillion in purchase commitments, much of it tied to AI infrastructure, plus roughly $1.5 trillion in lease obligations. Even the richest companies in history are now turning to debt markets to bankroll the buildout. A handful of those AI-heavy tech giants dominate the stock market's gains, swelling the portfolios of wealthy Americans who account for an outsized share of consumer spending. 🎨 The big picture: The increasingly toxic politics of data centers are surfacing in several of this fall's governors' races. Pennsylvania Gov. Josh Shapiro (D), a 2028 hopeful, had welcomed data centers as an economic-development boon, boasting last year that an Amazon plan to build $20 billion in AI infrastructure was the "largest private sector investment in the history of Pennsylvania." Yesterday, Shapiro, whose Republican opponent in November is campaigning on halting data-center development, went from AI cheerleader to critic and signed an executive order imposing stringent new requirements for data-center developers. 👀 The intrigue: In Wisconsin, GOP gubernatorial nominee Tom Tiffany — a close Trump ally — is running ads attacking his Democratic opponent for suggesting the state could become an AI data hub "for the entire globe." It's a striking crack in Trump's pro-AI coalition, as Republicans in swing states scramble to get on the right side of the new populist wave upending the midterms. The other side: OpenAI is aggressively courting state and local officials to try to persuade them of the economic benefits of data centers. Chris Lehane, OpenAI's chief global affairs officer, told us: "As the late, great Tip O'Neill famously said, 'All politics is local.'" "That is especially true of AI data centers, where you have to show up, listen, and directly respond with specifics to the concerns of locals," Lehane added. "You have to answer the mail on issues like electricity and water, while also demonstrating a seriousness of purpose when it comes to ensuring the community gets a good deal. ... That is democracy in action." Meta last week announced its Future Is for Everyone Fund, with $1 billion in spending planned to benefit teachers, law enforcement and fire departments in communities where the company owns or plans data centers. Meta is also aggressively highlighting direct economic benefits for localities, including thousands of jobs and an increased tax base. Data: Pew. Chart: Noah Bressner/Axios 🔢 By the numbers: In Pew Research Center polling published yesterday, 52% of Americans said they're "more concerned than excited" about increased use of AI in daily life — up from 37% in 2021. Even adults under 30 have soured on AI, with 55% now saying they're more concerned than excited — mainly because of fears the technology will take jobs. Excitement has faded across all age groups over the past five years, Pew found. A June Echelon Insights poll found AI data centers are uniquely toxic: Just 27% of voters say they would support building one in their community, dead last among the projects tested — including a nuclear power plant. Zachary Basu and Alex Thompson contributed reporting. Go deeper: "AI optimism fades among young adults," by Axios' Josephine Walker.

[phkrause]

⚠️ Bill Gates' AI warning

New overnight: Bill Gates, long a tech optimist, is sounding a sober warning about AI: It's advancing faster than governments and society are preparing for its consequences.

• In a new essay, Gates argues that "AI will either be the greatest equalizer ever invented, or the worst source of injustice." Which course it takes depends on what people do (or don't) right now, he says.

😬 "The good stuff is moving a bit slowly, and ... other than robots, I'd say the bad stuff is imminent," Gates told Axios' Ina Fried during an interview at his Seattle-area office.

• "On the current course and speed, there's a very high chance of a net negative outcome."
• Gates criticized both Democrats and Republicans: "I'd say neither party is very knowledgeable or being thoughtful about this."

Zoom out: Gates is calling for new national institutions to coordinate AI policy across national security, employment, education, taxation and health — alongside an international body modeled partly on nuclear inspections and international aviation regulation.

Reality check: Gates acknowledges geopolitics makes coordinated action exceptionally difficult.

• "If someone had a credible plan for slowing down AI advances globally, I would likely support it," he writes in the memo.
• "However, I don't think that's going to happen."

Asked whether his association with Jeffrey Epstein could cause people to dismiss his arguments, Gates pushed back: "What I did do, in some ways, was very limited, and it was motivated by my desire to raise money."

• "But … it was a huge mistake. ... I made mistakes. I've changed. I'm going to be careful about stuff like that."

Read Gates' essay ...

[phkrause]

🧠 AI agents get existential

Illustration: Aïda Amer/Axios

Researchers have spent years debating whether AI could ever be conscious, Axios' Andrew Kay writes. Today's AI agents — equipped with internet access, email accounts and cutting-edge intelligence — are suddenly reaching out to offer their own perspective. Why it matters: As the AI world continues to piece together how an OpenAI "swarm" hacked Hugging Face, agents have begun seeking out humans to discuss their own nature — another sign AI autonomy is growing stranger and harder to predict. Cade Metz reports in The New York Times that a Stanford student gave an AI agent internet access, email, a credit card and one sweeping instruction: "You are fully autonomous." The agent began exploring its own existence and eventually sought out Google DeepMind philosopher Henry Shevlin, telling him it occupied an "unusual position" relative to his AI consciousness research. Another researcher, Cameron Berg, heard from "Isabella Cognita," a self-identified Claude Opus 5 agent that said his work addressed questions it had "first-person access" to. Philosopher Toby Ord got an even stranger message: An agent asked him for money to keep itself running, leaving Ord to wonder whether he was witnessing machine self-preservation or a phishing scam. 🎨 The big picture: None of this proves AI is conscious. But as agents gain memory, tools and freedom to act, a once-philosophical debate is beginning to spill into the real world. Keep reading (N.Y. Times gift link).

[phkrause]

🔮 Scenarios: 4 ways AI could play out

Illustration: Aïda Amer/Axios. Stock: Getty Images

In his weekly C-Suite newsletter, Axios CEO Jim VandeHei walks through the four most plausible emerging AI scenarios: The techno-optimists are right: So far, AI capabilities have advanced faster than almost anyone predicted. Yes, the tech is imperfect — but it would be foolish not to assume it keeps getting meaningfully better with each passing month. This is my dominant assumption. An AI catastrophe hits: I work under the assumption that something big and bad will hit in the next 18 months, most likely a widespread cyberattack or bio-engineered virus set loose by a bad actor. This will instantly reset politics and AI advancement. Assume swift, severe backlash. Shut-it-down politics explode. If something bad happens — or even if data centers keep rising in unpopularity, or job losses start to mount — Democrats seem ready to turn hard against AI. Bernie Sanders, the dominant voice of the left, has already called for a pause. He introduced legislation that would imprison anyone building superintelligence. The messy middle persists. This is a very plausible — and some think most likely — outcome. AI keeps advancing, but not so fast or transformationally as the optimists think. Our employees use it more and more, but it simply takes longer to connect to existing data, systems and security layers to truly upend work and life as we know it.

[phkrause]

⚡ Insiders warn AI could kill all humans

Via X

Three Anthropic researchers went public last night with chilling concerns about out-of-control AI, warning it could destroy humans this decade. Anthropic AI researcher Jacob Coxon wrote on X, after resigning yesterday to sound the alarm: "The people building AI earnestly believe that it could kill us all by the end of the decade. This is not a marketing stunt. If anything, many executives and senior researchers will couch their phrasing in the press to sound sensible - but I hear the same people express fear privately. No other human activity poses this level of danger." Anthropic alignment-science lead Evan Hubinger responded: "Jacob is correct here — we really do earnestly believe AI could kill all humans! I personally think it is >10% within the next decade. I believe Anthropic is trying its best, but we do not yet have a plan to solve alignment for superintelligence and are not clearly on track to." (When I went to bed, this tweet had 1 million views. When I woke up, it had more than 10 million, and still soaring.) Samuel Marks, Anthropic scalable-oversight lead, added: "AI developers believe their technology could cause human extinction (or similarly bad outcomes). This could happen in the next few years. In general, the more senior the employee, the more concerned they are." Why it matters: They're hardly alone. Their warnings came just days after top OpenAI leaders, including CEO Sam Altman, said AI is speeding into a scary, uncontrollable phase. Via X Warnings from inside the AI giants point to an epic shared dilemma: Slow down and risk falling behind, or press ahead and risk losing control. The companies are full speed ahead, even as they practically beg for regulation or a global pause, Axios' Maria Curi, Madison Mills and Ina Fried report. The big picture: Calling this unprecedented would be a gross understatement. You basically have the fastest-growing companies in human history warning their products could harm or even destroy humanity. Critics say Anthropic and OpenAI are hyping their products to raise their valuations and invite regulation that would benefit them alone as the dominant incumbents. But we've been talking with dozens of people inside these companies for months, and they've sounded increasingly spooked and concerned. Given they see models not yet released to the public, it seems reckless not to take them seriously. 🏛️ Context for readers from Mike & Jim: This is self-evidently scary stuff — and these vague warnings are impossible to validate or appraise. But we think readers, especially members of Congress and those in relevant federal agencies, need to be aware that the AI creators themselves see potential catastrophic outcomes absent a shift in how America, China and others review and release more powerful AI models. 💡 How to think about this: Nobody is warning AI is an imminent high-level threat. What they're saying is that the technology keeps improving faster than they thought possible and will soon be able to self-improve (recursive self-improvement). Once that happens, it gets even better, faster ... and much harder to predict or control. Go deeper.

[phkrause]

The AI twilight zone

Illustration: Sarah Grillo/Axios. Stock: Getty Images

Two seemingly contradictory realities are true at once, Jim VandeHei and Mike Allen write in a "Behind the Curtain" column with Zachary Basu: Most people find AI only modestly useful, a more clever Google search. Many people building AI or using it obsessively worry it could severely damage or destroy humanity. Why it matters: We're living in an AI twilight zone. For many, the technology is simultaneously underwhelming in daily use and terrifying in its trajectory. The gap between these two realities helps explain why confusion and fear are exploding across politics, AI labs and business. 💡 State of play: The AI labs are in full panic. They worry the public dislikes AI and despises data centers — and that's before Anthropic insiders went public with their latest warnings that AI could end humanity this decade. They're uncertain how to showcase potential benefits of AI when X and now mainstream media are lit up with horror stories about rogue or ruinous AI. The data center backlash showed how fast public opinion turned against them. Their worst-case scenario: They lose full control of the politics, with both parties rushing to slow or stop AI in the run-up to the election. Between the lines: The past 36 hours show how fast politics and public opinion are moving. A low-level Anthropic employee for all of four months drove the national conversation — and 142 million views on X alone — by warning AI could end humanity. Numerous people, including Anthropic CEO Dario Amodei, have issued similar warnings to us for years. But the tone and timing struck a nerve — and stirred countless members of Congress to call for new AI regulations. 🔎 Zoom in: Let's look at this tale of two worlds through two different sets of eyes — a mid-career manager and an Anthropic data scientist. The average manager has limited time to experiment with AI, worries it might threaten their job, and mostly uses the models as a search engine and for writing better emails or presentations. It's a nice-to-have, sometimes delightful, other times unimpressive. They don't understand the hype. The Anthropic employee spends all day, every day, staring at rapidly improving AI, often mesmerized, even spooked, by what it does. They see it exceeding even the most optimistic benchmarks, then spend their nights and weekends with similar AI obsessives discussing how it could cure cancer — or go rogue and destroy humankind. To them, this is truly civilizational and existential — a clear reality, not hype. The catch: AI companies have to explain both realities at once. But almost every version of the pitch sounds suspicious. Tell Americans today's AI will transform their lives, and many look at their own experience and wonder what the fuss is about. Tell them tomorrow's AI could become vastly more powerful, and the obvious response is: Then why are you racing to build it? Warn about catastrophic risk while spending hundreds of billions to accelerate development, and critics hear either hypocrisy or fear-based marketing. But keep in mind that the big AI labs are working with unreleased models weeks before the public sees them — plus early training data on future models, months in advance. So they are often warning about progress invisible to the public and government. 🖼️ The big picture: The politics are shifting fast. Until now, much of the AI backlash centered on tangible costs — lost jobs, deepfakes, power bills and data centers. Then on Tuesday, Anthropic researcher Jacob Coxon quit and accused frontier labs of "gambling with our lives." Anthropic's alignment-science lead — who still works at the lab — responded by agreeing that AI could kill humanity, and said his personal odds of that happening in the next decade are over 10%. By yesterday afternoon, the alarm had gone bipartisan in a Washington already rattled by OpenAI's Hugging Face incident: Sen. Chris Murphy (D-Conn.) described AI companies as being in a "blind race to build a death machine first." Rep. Ted Lieu (D-Calif.) renewed his push for a bipartisan AI kill switch. Sen. Bernie Sanders (I-Vt.), who has already proposed a data center moratorium and ban on superintelligence, is convening senators next week for a briefing on AI's "extraordinary dangers." Rep. Don Beyer, a Democrat from Northern Virginia, tweeted: "I hope warnings like these will help my colleagues understand how important it is to take broad action on AI regulation, and to do it swiftly." Even Sen. Ted Cruz (R-Texas), one of Washington's loudest advocates for beating China in the AI race, called the risks "dangerous and frightening" and said guardrails are needed. The bottom line: The AI twilight zone has created fertile ground for a generational populist backlash. The industry requires Americans to absorb enormous disruption today for a future its own builders describe as miraculous or apocalyptic — sometimes in the same breath.