🚨New York signed the RAISE Act requiring AI developers to report safety incidents within 72 hours and publish safety protocols, becoming the second state to enact major AI safety legislation despite Trump's executive order attempting to block state regulation—proving states aren't backing down from regulating AI even under federal pressure.

Unlock a focused set of AI strategies built to streamline your work and maximize impact. This guide delivers the practical tactics and tools marketers need to start seeing results right away:

Stay ahead of the curve with these top strategies AI helped develop for marketers, built for real-world results.

Download the Free Report

We've all seen the headlines: AI data centers are draining the planet dry, guzzling billions of liters of water while communities face droughts. It's a compelling narrative—silicon valley literally sucking our future dry.

But what if the real story is completely different?

One analyst has done the math, and the numbers tell a surprising tale. According to their calculations, even at the high end of estimates, global AI data centers might use up to 765 billion liters of water in 2025.

Sounds like a lot, right?

Here's the twist: U.S. golf courses alone used around 2.85 trillion liters for irrigation back in the mid-2000s. That's nearly 4x more water for keeping fairways green than powering the entire global AI infrastructure.

And it gets more dramatic...

The actual scale:

The United States uses more water in a single day than all AI data centers are projected to use in an entire year. We're talking about a resource that represents a tiny fraction of global consumption—far smaller than agriculture, industrial manufacturing, or even recreational activities like golf.

The analysis doesn't dismiss water concerns entirely. Placing data centers (or golf courses, for that matter) in drought-prone regions like Arizona or Nevada? That's genuinely problematic and requires smarter planning.

But globally, AI's water footprint is minor. And here's the part people forget: water isn't destroyed—it cycles back through evaporation and rainfall. It's not disappearing into a digital void.

So why the panic?

The water narrative might be distracting us from AI's real risks—the ones that actually matter for humanity's future.

The economic risk: What happens to human work when robots and AI can handle most physical and mental tasks? We're not talking about automating a few jobs here and there. We're talking about fundamental questions of purpose, income, and societal structure when human labor becomes optional.

The alignment risk: How do we ensure increasingly capable AI systems act in humanity's best interests? As these systems become more powerful, the stakes grow exponentially. One misaligned objective function, one unintended optimization, and we could face consequences far more severe than water scarcity.

The author's point is clear: we're obsessing over AI's relatively tiny water use while ignoring the existential questions it poses. It's like worrying about your laptop's battery drain while your house is on fire.

What's next

As AI capabilities accelerate, the conversation needs to shift. Water management for data centers? Sure, that deserves attention—especially at the local level. But the bigger questions are economic transformation and AI safety.

How do we build an economy where humans thrive alongside increasingly capable AI? How do we ensure AI systems remain aligned with human values as they become more autonomous and powerful?

These aren't abstract philosophical questions anymore. They're urgent policy and technical challenges that will define the next decade.

BTW: The average American uses about 300-400 liters of water per day. At that rate, just 5-6 million Americans use more water annually than the entire global AI infrastructure. That's roughly the population of Wisconsin. So next time someone says AI is causing a water crisis, you can tell them Wisconsin is the real culprit. (We kid, Wisconsin. You're doing fine.)

Have an ad concept ready but don't want to deal with expensive shoots or stock footage? ScriptKit lets you generate, curate, and edit AI ads in one platform.

What ScriptKit gives you

Give ScriptKit a shot — go from concept to finished AI ad without wrangling teams or gear.

The Recap: iRobot filed for Chapter 11 bankruptcy last Sunday after Amazon abandoned its $1.7 billion acquisition following an 18-month FTC investigation, marking the end of an era for the Roomba maker that had sold over 50 million robots since 2002 and survived 35 years of near-death experiences. During his deposition, founder Colin Angle walked the halls of the FTC and noticed examiners had "printouts of deals blocked, like trophies" on their office doors. The collapse exposes how regulatory enforcement can reshape entire industries by making acquisition exits unreliable, forcing hardware startups to compete on capital efficiency rather than consolidation strategies.

Unpacked:

Bottom line: When regulators kill deals, they don't just block individual transactions—they fundamentally restructure how entire categories of companies can scale and exit. The FTC celebrated blocking the Amazon deal like a "trophy" on their wall, but the actual outcome was a 35-year-old American robotics pioneer going bankrupt and selling to Chinese interests. Angle's observation about FTC examiners displaying blocked deals as achievements captures the perverse incentives: the agency optimizes for stopping consolidation rather than preserving American technological leadership or consumer welfare.

TikTok Shop launched digital gift cards ($10-$500) during the holiday season to compete directly with Amazon and eBay, embedding commerce into social engagement rather than forcing users to leave the platform—executing the WeChat playbook that made Asian super-apps dominant.

The launch comes after TikTok Shop recorded over $500 million in U.S. sales during Black Friday/Cyber Monday, even as the app faces a potential ban with Trump extending the deadline to January 23, 2026, leaving its future hanging in the balance.

Waymo suspended robotaxi service in San Francisco after vehicles stalled during a regional blackout, exposing the dependency risk of city-scale autonomous networks on centralized infrastructure and raising questions about real-world deployment resilience when power fails.

OpenAI allows users to directly adjust ChatGPT's warmth and enthusiasm, acknowledging that LLMs as commodities require personalization layers to create defensible user preference lock-in—a strategic pivot from raw capability competition to behavioral customization.

Disney's robotic Olaf character development signals how entertainment companies can defensibly differentiate through proprietary hardware that competitors cannot quickly replicate, as theme parks become experiential technology platforms where physical robotics creates competitive moats.

Italy's antitrust authority fined Apple for anti-competitive practices, part of coordinated global enforcement targeting the same dominant platform behaviors simultaneously across jurisdictions—showing tech leaders can no longer rely on regulatory arbitrage or delayed international compliance.

EU waters down 2035 EV goals after pressure from incumbent automakers, revealing that when environmental commitments clash with entrenched industries' political power, regulations designed to force tech disruption get rewritten—policy alone cannot overcome business interests.

Local LLMs undermine cloud AI provider economics as open-source models become production-ready and on-device inference improves, threatening the unit economics of companies built on managed cloud AI services when customers can run models locally.

Best of AI™ Team

Was this email forwarded to you? Sign up here.