But as this happens, it’s crucial to keep the denominator in mind. Since 2020, Waymo has reported roughly 60 crashes serious enough to trigger an airbag or cause an injury. But those crashes occurred over more than 50 million miles of driverless operations. If you randomly selected 50 million miles of human driving—that’s roughly 70 lifetimes behind the wheel—you would likely see far more serious crashes than Waymo has experienced to date.

Federal regulations require Waymo to report all significant crashes, whether or not the Waymo vehicle was at fault—indeed, whether or not the Waymo is even moving at the time of the crash. I’ve spent the last few days poring over Waymo’s crash reports from the last nine months. Let’s dig in.

Earlier this month, OpenAI released a new product called Deep Research. Based on a variant of the (still unreleased) o3 reasoning model, Deep Research can think for even longer than conventional reasoning models—up to 30 minutes for the hardest questions. And crucially, it can search the web, allowing it to gather information about topics that are too new or obscure to be well covered in its training data.

The coming AI speedup
The success of Deep Research also suggests that there’s a lot of room to improve AI models using “self play.” The big insight of o1 was that allowing a model to “think” for longer leads to better answers. OpenAI’s Deep Research demonstrates that this is true for a wide range of fields beyond math and computer programming.

And this suggests there’s a lot of room for these models to “teach themselves” to get better at a wide range of cognitive tasks. A company like OpenAI or Google can generate training data by having a model “think about” a question for a long time. Once it has the right answer, it can use the answer—and the associated thinking tokens—to train the next generation of reasoning models.

Explaining actual AI products is the core of Timothy B. Lee’s excellent Substack newsletter Understanding AI. Lee, a journalist (and occasional Reason contributor), refreshingly covers AI like a normal newsworthy subject. His articles include a nice range of original reporting on the companies and nonprofits producing AI, service journalism on how ChatGPT compares to Gemini, even-handed analysis of the legal and regulatory questions AI has inevitably provoked, and explainer articles on what even is a large language model.

The biggest takeaway is that, for all the boosterism and doomerism, AI will be a normal-ish technology that will have normal-ish impacts on the world. One of Lee’s best entries is a deep dive into how AI has affected one of the industries where it already predominates: language translation. Turns out that prices for translation have fallen and companies consume more translation services. That’s an unambiguous win for consumers.