Artificial Intelligencer-AI's coding war hits Super Bowl ads

By Krystal Hu

Feb 11 (Reuters) - (Artificial Intelligencer is published every Wednesday. Think your friend or colleague should know about us? They can subscribe to our newsletter here or email me to share any thoughts.)

If you're anything like me, the moment you feel something unusual in your body — a strange symptom, a nagging pain — your instinct these days is to ask Dr. ChatGPT or Dr. Gemini.

We’re not alone. For years, we’ve turned to search engines for quick medical answers. Now, AI chatbots promise something more: personalized responses, tailored to the details we share.

According to OpenAI, one in four users submits a healthcare-related prompt every week, and more than 40 million people turn to ChatGPT daily with medical questions. The opportunity is so large that OpenAI is preparing a ChatGPT for Healthcare, designed to support the “evidence-based reasoning required in real patient care.”

But the “AI doctor in your pocket” vision isn’t all progress and promise. In a new investigation, my Reuters colleagues uncovered cases where heavy reliance on AI for medical advice caused unnecessary anxiety. At the same time, a growing number of mobile apps on Apple and Google app stores claim to use AI to guide patients through medical issues — even though they aren’t supposed to provide diagnoses. One was removed after Reuters raised concerns.

In today’s issue, we also unpack what Super Bowl AI ads reveal about the latest battlefield in the race to win paying customers, and what earnings season told us about how big tech is budgeting for AI spending ahead. Scroll on.

Our latest reporting in tech & AI

Exclusive - ByteDance developing AI chip, in manufacturing talks with Samsung

Investigation - As AI enters the operating room, reports arise of botched surgeries and misidentified body parts

Two co-founders of Elon Musk's xAI resign, joining exodus

Anthropic buys Super Bowl ads to slap OpenAI

Alphabet sells rare 100-year bond to fund AI expansion as spending surges

China enlists AI to sniff out corruption in public bidding

Super Bowl ads show AI’s renewed battleground

Bad Bunny might be the star of the halftime show. But the ads in between all the breaks are showing where the money is these days — mostly in AI, and especially coding tools.

Between commercials for snacks and weight-loss drugs, AI companies made a different kind of pitch. Not just smarter assistants. Not just creative tools. But something more fundamental: that software itself is now the top battleground in AI. For all the talk of productivity and transformation, the money is increasingly flowing into one use case: helping everyone write code.

OpenAI used one of the Super Bowl’s most expensive ad slots to spotlight its AI coding agent, Codex. The 60-second commercial — with the tagline “You can just build things” — was designed to position Codex not as an app, but as a creative enabler.

The spot leans into a broader theme: that everyday people can use AI to turn ideas into real products. The goal was to inspire viewers to think not just about what AI can answer, but what it can help them create, from websites to tracking apps.

From helping coders work faster and non-coders to kick off building software, coding has become AI’s ultimate “high economic value” task. Engineers are expensive. Output is measurable. And models are particularly strong at reasoning through structured problems with verifiable answers. GitHub Copilot proved early that AI-assisted coding wasn’t just hype — it could slot directly into daily workflow.

Anthropic leaned into this aggressively early on. By focusing model training on coding, it built a reputation — and a revenue base — around developer APIs that perform best for coding tasks, and capitalized on it by launching its own coding product, Claude Code. That focus helped push its latest valuation to around $350 billion.

OpenAI knows this is where the fight is and now is trying to invest in both R&D and marketing to regain the ground. Its newest model, GPT-5.3-Codex, currently tops many benchmarks evaluating the model’s ability, from coding to reasoning.

But benchmarks are one thing; developer switching is another. Meanwhile, Google is expanding its own coding push, strengthened by its acquisition of Windsurf. Chinese open-source players — from DeepSeek to Qwen to Kimi — are rapidly improving, lowering barriers for coding startups to compete.

For now, Anthropic holds the edge. Investors credit its close focus on model coding instead of general capabilities, tight feedback loop with enterprise users, and infrastructure stack that prioritizes fast inference and reliability.

This isn’t just a model competition. It’s a revenue race. Whoever wins the coding war could build AI’s first truly durable business model, because software creation itself is what people are willing to pay a price for.

Chart of the week

Despite a volatile week in markets, one theme from Big Tech’s latest earnings was consistent: AI spending this year is accelerating, not retreating. The chart shows just how large the commitments have become. Amazon, once seen as an AI laggard, now plans to spend roughly $200 billion on AI-related investments in 2026, leading the pack. Collectively, Big Tech is expected to pour more than $600 billion into AI next year, extending what has already been a historic ramp-up in capital expenditure.

Much of the spending is flowing into servers, networking equipment and advanced chips, positioning companies like Nvidia as major beneficiaries. What’s new here is the capital isn’t coming from Big Tech’s operating cash alone. After initially relying heavily on their own balance sheets, they are increasingly tapping the bond market to sustain the pace. Bank of America estimates that five major tech firms could borrow around $140 billion annually over the next three years to fund AI expansion. Wall Street expects spending to plateau by 2027 — but for now, the AI arms race remains a capital-intensive sprint, even as investors raise eyebrows over how quickly it will translate into durable profits.

(Reporting by Krystal Hu; Editing by Lisa Shumaker and Rosalba O'Brien)

((krystal.hu@thomsonreuters.com, +1 917-691-1815))