AI Is Already Here: Her Radiologist Used It on Her Scans

On May 12, 2026 — the same day this article publishes — Joanna Stern released I Am Not a Robot, a book that started as a dare and ended as a quiet verdict on AI automation: the tools are already good enough, already deployed, and already running in places most people never consented to. She spent 12 months integrating AI into every corner of her personal and professional life, testing dozens of tools, and letting technology disrupt family dinners and her editorial team alike. What surprised her most had nothing to do with chatbots.

Her own radiologist had been quietly using AI to analyze her medical scans for an entire year — without ever telling her. No announcement. No consent form. Just AI, already running.

From The Wall Street Journal to Living Inside the Machine

Stern is one of the most credentialed tech journalists in the US. She co-founded The Verge, spent years as The Wall Street Journal's lead technology columnist, and built her career on rigorous skepticism over easy enthusiasm. That pedigree makes her 12-month experiment unusually credible — and her conclusions unusually uncomfortable.

For the book, Stern didn't review AI tools from a distance. She surrendered to them, season by season. Her wife and children participated in product trials. She replaced a human researcher on her team with an AI agent (an automated assistant capable of completing multi-step tasks like research, scheduling, and document summarization), then invited that displaced researcher back for an on-camera interview about the experience. The book documents all of it — the wins, the failures, and the places where AI showed up without being invited.

After the year ended, Stern left institutional media entirely to launch New Things, an independent media company built in partnership with NBC. Her bet: audiences are exhausted by hype and hungry for honest AI assessment. The book is the first product of that conviction.

The Radiologist Reveal — and the Bigger AI Automation Pattern Behind It

The most striking discovery in Stern's year wasn't a product failure. It was a quiet, invisible success she'd never consented to.

Her radiologist had been running AI-assisted analysis on her medical imaging (software that detects anomalies in X-rays and scans faster and often more accurately than unaided human review) for a full 12 months before Stern found out. The AI was embedded in the practice's existing infrastructure — working silently, working well, and never appearing on any consent form.

This, Stern argues, is the actual story of AI adoption in 2026. Not chatbots. Not humanoid robots. Invisible integration:

• Healthcare: AI reads scans, flags anomalies, and routes records — usually with no patient-facing notice
• Transportation: Waymo's self-driving taxis (geographically limited to select US cities, currently expensive, subscription-based) show that functional, non-chatbot AI exists at scale
• Search: Google's AI Overviews now appear above organic results for hundreds of millions of users who never opted in
• Social media: Instagram's feed is AI-curated for roughly 2 billion users, reshaping what they see every day without a settings prompt

"There are going to still be ways that AI affects your life regardless of whether you want it to." — Joanna Stern, I Am Not a Robot

The book includes a full chapter on Waymo as a case study: AI that works, but that consumers didn't actively choose — it arrived via city permits and corporate partnerships, not an App Store download. Healthcare follows the same pattern. These aren't edge cases. They're the dominant adoption model.

AEI vs. AGI — Why 'Artificial Enough Intelligence' Wins in 2026

Near the end of her 12-month experiment, Stern coined a phrase she believes better captures where AI actually stands: AEI — Artificial Enough Intelligence.

The concept is a direct challenge to the industry's obsession with AGI (Artificial General Intelligence — the theoretical threshold at which AI matches or surpasses all human cognitive abilities across every domain). Stern's argument: nobody needs that. The tools already available in 2026 are sufficient for most real-world tasks. The bottleneck isn't AI capability — it's poor application of existing tools.

"We don't need AGI. A lot of these tools that we already have are good enough and they just have to be applied better." — Joanna Stern

Her breakdown of AEI in practice — after testing approximately 3–4 major consumer AI platforms (ChatGPT, Gemini, Claude) plus dozens of specialty tools over 12 months:

• ✅ Already works: AI-assisted medical imaging, Slack workflow bots, document summarization, self-driving in geo-fenced areas
• ❌ Not ready: Humanoid robots ("definitely not ready, and they might not be for a very long time" — Stern's words after interviewing multiple teams building them); consumer chatbot interfaces that haven't materially changed in roughly 4 years
• ⚠️ Watch this space: Wearable AI — Stern identifies this as the potential "killer app" (the first AI category consumers might actually choose with enthusiasm, rather than simply accept)

The AI Tools Interface Problem Nobody in the Industry Wants to Say Out Loud

One of Stern's sharpest observations targets AI product design specifically. Since ChatGPT launched approximately 4 years ago, the core consumer interaction model — type a question, read an answer — has barely changed. The only meaningful interface improvement she credits across the entire industry: voice mode, which arrived years after launch and remains a niche feature for most users.

Nilay Patel, host of The Decoder podcast (The Verge's flagship long-form interview show, which produced the primary source conversation for this article), said it directly:

"I don't think consumer AI products are very good. I don't think there's a great consumer AI product, and I think a ton of the angst we hear about AI is a reflection of that."

Stern compares this stagnation to the iPhone era, when Apple products were "obviously great" and consumers chose them with genuine enthusiasm. Today's AI adoption looks different at ground level:

• Google Search inserts AI Overviews before traditional results — users encounter AI whether they seek it or not
• Instagram's recommendation algorithm has been AI-powered for years, invisible to most of its 2 billion users
• Healthcare AI (like Stern's radiologist's imaging tools) operates below the surface of existing clinical infrastructure

The pattern: AI's biggest wins aren't products people downloaded. They're systems people never noticed. Consumer AI, by contrast — the chatbots, the AI gadgets, the humanoid robots — is fighting for attention it hasn't yet earned through utility.

Her AI Automation Team Setup — and What You Can Copy Right Now

Stern's business workflow offers a concrete template for non-technical teams who want practical AI without enterprise pricing:

• Hardware: A Mac Mini (Apple's compact $499–$799 desktop — no GPU cluster, no cloud server contract required)
• Workflow layer: A Slack bot (a custom automated assistant inside the team's messaging app) connected to an AI agent that handles multi-step tasks like research routing, document summarization, and scheduling
• Guiding principle: Eliminate "busywork" (repetitive, low-creativity tasks like reformatting documents, pulling basic research, and writing meeting recaps) first; actively protect creative roles like video editing and narrative storytelling

The important nuance: Stern replaced her human researcher role with an AI tool, then rehired that person for higher-value creative work the AI couldn't replicate. Her goal was never headcount reduction — it was task reallocation toward work that actually requires humans.

If you're a marketer, designer, or office worker trying to figure out where to start, Stern's framework is simple: identify the three most repetitive tasks you completed last week. Those are your AEI candidates — the work that's already "artificial enough" to hand off. You can explore specific tool recommendations in our step-by-step AI automation guides, built for non-technical readers ready to act today.

Stern's book won't tell you which chatbot ranked highest in a benchmark. What it will tell you is more useful: the AI most actively reshaping your life isn't the app you downloaded. It's probably already operating in your doctor's imaging suite — and has been for longer than you know.

Related Content — Get Started with AI Automation | Guides | More News