But Google's global search dominance makes it the most important player, and the company has already rolled out AI Overviews to more than a billion people worldwide. The result is searches that feel more like conversations. Google and OpenAI both report that people interact differently with generative search--they ask longer questions and pose more follow-ups. This new application of AI has serious implications for online advertising and (gulp) media. Because these search products often summarize information from online news stories and articles in their responses, concerns abound that generative search results will leave little reason for people to click through to the original sources, depriving those websites of potential ad revenue.

Most big tech firms now boast fun-size versions of their flagship models for this purpose: OpenAI offers both GPT-4o and GPT-4o mini; Google DeepMind has Gemini Ultra and Gemini Nano; and Anthropic's Claude 3 comes in three flavors: outsize Opus, midsize Sonnet, and tiny Haiku. Microsoft is pioneering a range of small language models called Phi. A growing number of smaller companies offer small models as well. The AI startup Writer claims that its latest language model matches the performance of the largest top-tier models on many key metrics despite in some cases having just a 20th as many parameters (the values that get calculated during training and determine how a model behaves). Smaller models are more efficient, making them quicker to train and run.

Now, roboticists have made major breakthroughs in that pursuit. One was figuring out how to combine different sorts of data and then make it all useful and legible to a robot. Take washing dishes as an example. You can collect data from someone washing dishes while wearing sensors. Then you can combine that with teleoperation data from a human doing the same task with robotic arms.