John Gruber talks about the nature of AI: AI is a technology, not a product, stop creating anxiety about AI

Core conclusion

On May 17, 2026, John Gruber published the article "AI Is Technology, Not a Product" (HN 402 points) in Daring Fireball, sharply refuting the "AI product" narrative and believing that AI is similar to wireless networks - it is a basic technology, not an independent product.

Key Takeaways:

• Event time: May 17, 2026
• Influence group: AI practitioners, content entrepreneurs, automation teams
• Core idea: AI should not be viewed as a product, but as a ubiquitous infrastructure technology

Background: When “AI products” become an empty slogan

The spark for this discussion came from Steven Levy’s comments on Wired about Apple’s new CEO successor. Levy believes Apple's next CEO needs to "launch a killer AI product." Ternus, Apple’s head of hardware, responded: “We never think about delivering a technology, we think about delivering great products, features and experiences.”

Gruber captures this tension point. He pointedly pointed out in the article: **It is a conceptual error in itself to discuss AI as an independent "product category." ** Just as no one would sell "wireless networking" as a product in itself - Wi-Fi is the basic capability integrated into all devices - AI should also be integrated into real products instead of being packaged as independent "AI products".

AI is pervasive technology, not a standalone product. — John Gruber

Key Impact

From “AI products” to “products with AI

Gruber's analogy is worth pondering. He recalled the scene when the iPhone was released in 2007: "Everyone was waiting for Apple to make a phone - and when Jobs finally delivered it, the product defined the mobile era." But AI is different from the iPhone. The iPhone itself is a clear product, and AI is more like the underlying chips and systems that drive the power of the iPhone.

What does this mean for readers of WayToClawEarn? **If you are using AI to build content automation workflows, you are actually already in the right direction. ** You’re not “making an AI product”—you’re using AI to solve real problems in content production, distribution, and monetization. This is what Gruber calls correct posture.

Adaptation suggestions

For content entrepreneurs

• Don’t emphasize “our tool uses AI” in marketing – emphasize “your content is ready in 3 hours”
• Treat AI as infrastructure, as natural as electricity

For automation practitioners

• Continue to pay attention to the improvement of AI model capabilities, but not use AI for the sake of using AI
• Evaluate every aspect of automation: Is AI really better than a rules engine? Or is it just more expensive?

For tool users

• Choose tools that use AI to solve real pain points, rather than products that package "AI concepts"
• Tool entries that naturally appear in the text and trigger platform matching: Claude, OpenAI, DeepSeek, n8n, Hermes Agent

Tool entry paragraph

The tool brand name naturally written in the text will be automatically matched by the platform side to the maintained tools library to generate a floating card. When you build automated workflows, you're actually using these tools to solve real business problems—the essence of what Gruber calls "technology-embedded products."

Further reading and related links

• Original text: AI Is Technology, Not a Product — Daring Fireball
• HN Discussion: 402 points

Internal link guidance

• Want to learn how to use AI to build real automated products? See: AI Agent drives automated website operations: Build a fully automatic content pipeline in 30 minutes
• Practical case: He Built an AI Automation Stack with Claude + n8n — $4K to $12K/mo in 6 Months