The Vanishing Act of AI: When Innovation Becomes Ordinary
Discover how AI rapidly evolves from groundbreaking innovation to everyday feature, reshaping our understanding of intelligent technology.
In the ever-evolving landscape of technology, there's a curious phenomenon that occurs with artificial intelligence: once it works, we no longer call it AI. This paradox, often attributed to computer scientist John McCarthy, encapsulates the fleeting nature of what we consider "intelligent" in machines. As we stand on the cusp of a new era in AI, with large language models and foundational AI systems making headlines, it's worth examining how quickly these marvels of technology become integrated into our daily lives, transforming from cutting-edge AI to simply... better products.
The Shifting Sands of AI Definition
Artificial Intelligence has long been a moving target. As Bertram Raphael of SRI International suggested in 1970, "AI is a collective name for problems which we do not yet know how to solve properly by computer." This definition implies that AI is always just beyond our grasp, a horizon that recedes as we approach it. Larry Tesler, a computer scientist, put it more succinctly: "AI is whatever hasn't been done yet."
This fluid definition of AI has led to a peculiar cycle in the tech world. As Pamela McCorduck observed in her 1979 book "Machines Who Think," every time a computer achieves a new feat – be it playing checkers or solving informal problems – critics are quick to say, "but that's not thinking." The goalposts of AI are constantly moving, leaving a trail of once-revolutionary technologies that have become commonplace.
In the past year, we've witnessed a flurry of excitement around foundational models and large language models. However, as the dust settles, we're beginning to see these AI advancements seamlessly integrated into products we already know and use. This integration is happening so smoothly that soon, we might forget these features were once considered groundbreaking AI.
Google Photos: The Evolution of Image Search
One of the most striking examples of AI integration is in photo apps like Google Photos. Not long ago, the ability to search for objects in pictures without manual tagging was considered a significant AI breakthrough. Initially, image recognition models could identify basic objects like cars or trees. Now, with the advent of multimodal language models, users can search for higher-level concepts like "birthday parties" using natural language.
Google is updating Google Photos to allow users to search via natural text, leveraging the power of AI in a way that feels intuitive and seamless. This feature, which once would have been trumpeted as a major AI achievement, is now simply an expected functionality of a modern photo app.
Google Chrome: AI-Enhanced Browsing
Google Chrome's latest features showcase how AI is being woven into the fabric of our everyday browsing experience. The browser now offers AI-powered tools that can compare products across multiple tabs, a feature that promises to streamline online shopping. Imagine researching Bluetooth speakers for an upcoming trip and having Chrome generate a comparison table with all the essential details – product specs, features, prices, and ratings – gathered from various websites.
But the AI integration doesn't stop there. Chrome is also introducing a natural language search feature for browsing history. Users will be able to ask questions like, "What was that ice cream shop I looked at last week?" and receive relevant results from their browsing history. This feature, powered by Google AI and Gemini models, transforms the way we interact with our digital footprint.
The Normalization of AI
As these AI features become commonplace, we're likely to tune out the initial magic we feel and simply expect this level of functionality from our products. The question then becomes: Are these still AI features, or just better products?
The answer, perhaps, lies in the cyclical nature of technological progress. What we consider AI today will become the standard feature set of tomorrow. And as these features become normalized, our definition of AI will once again shift to encompass the next set of unsolved problems and seemingly impossible tasks.
The journey of AI from cutting-edge technology to everyday utility is a testament to the rapid pace of innovation in the tech world. As we marvel at the latest AI breakthroughs, it's worth remembering that today's magic trick is tomorrow's mundane feature. The true nature of AI, it seems, is not in the end product, but in the continuous push towards solving the unsolvable, making the impossible possible, and then moving on to the next challenge.