Thoughts on GoogleIn 2023, Google Search generated $175 billion in revenue. The next closest unit was Google Cloud with $33 billion (YouTube generated $31.5 billion, in case you were wondering). That’s a pretty big step down from your top money-maker to second place. And that’s why we’ve seen a big change in how Google is operating. For over two decades, we’ve been trained to think of search as the way to find important information. Need to know where the closest supermarket is? Google it. What time is the playoff game? Google it. What was the name of the drummer in Aerosmith? Google it. All Google had to do was keep the servers running and the money printing machine would continue to print money. Then ChatGPT came onto the scene. Enter ChatGPT. While ChatGPT opened our eyes to a lot of things, the most important from Google’s perspective is that you can now find information without search. Instead of typing that question into a Google search box, you can ask ChatGPT. Instead of being met with a page of links you need to investigate, the first four of which you scroll by because they’re ads, you get a concise answer from ChatGPT. To get a sense for the fear this struck in Google, it helps to understand that ChatGPT was publicly released in November 2022. In January 2023, Google laid off 12,000 employees to “focus on AI” and then laid off hundreds more in 2024. Google understands this deeply: AI chatbots are coming for search. They are building their cash stockpile for a cold winter that could come sooner than anyone thinks. Search’s days may be numbered. Google is consistently rolling out announcements about Gemini, their own model. Most of the time, they seem to be significantly behind the work at OpenAI, but at this point, that might not matter. The only question now is whether people will prefer asking a question to an AI or searching. Even if a small percentage choose the AI experience, that could cut into Google’s search advertising significantly. Even if Gemini proves to be superior to GPT in the long run, that still doesn’t guarantee Google anything. Summarizing what people are looking for doesn’t lend itself to advertising. To date, there is still no strategy for effectively monetizing AI. AI startups aren’t making money. So even if Google wins the AI battle, it may lose the monetization war. When you see people bemoaning that Google isn’t the friendly, open company they once were, and that they’re putting profits before everything else, this is why. Google correctly identified AI as an existential threat to their money-printing machine (search advertising). Now, they are doing everything they can to convince shareholders that they are ready to face this challenge. But are they? Key Takeaways
Stuff I've Enjoyed this Month📚 Don't Shoot the Dog by Karen Pryor 📚 Writing for Busy Readers by Todd Rogers and Jessica Lasky-Fink 🎬 The Latest in Web UI by Chrome Developers 🎬 JavaScript Visualized - Event Loop, Web APIs, (Micro)task Queue by Lydia Hallie 📝The problem with new URL(), and how URL.parse() fixes that by Kilian Valkhof What I'm Working On🏠 Real Estate: It has been a slow month for my rental properties, which is a good thing! My home, on the other hand, received an upgrade in the form of a resurfaced back deck. I replaced old rotting wood with a composite deckboard. Follow my Instagram for real estate photos. 📝 Writing: I wrote a blog post explaining how someone temporarily took control of one of my subdomains. This was quite the learning experience. 💻 ESLint: One of the big complaints after ESLint v9 came out was that the migration was difficult. I spent a lot of time hunting down the problems people were having, resulting in a pull request to Rush Stack, a new migration tool, and submitting pull requests to several projects who requested help. |
A once-per-month newsletter discussing topics important to senior-level software engineers, with a particular focus on frontend technology and leadership.
Code portability If you’ve been following me recently, you’ve probably seen me mention Bredbox, my new application for saving bookmarks. This is the first time in years I’ve built a web application, so I put a lot of thought into the process. My philosophy is to leave as many options open as possible in case I need to change direction later. That’s why I recommend targeting Node.js for JavaScript applications, even if you plan to run them on Deno or Bun. Both Deno and Bun are largely...
The return of tech specs When I was studying computer science as an undergraduate from 1996 to 2000, I took a course called Software Engineering that focused on the formal process of developing software. We learned to write a functional requirements document describing what the application should do, followed by a technical specification outlining how the software would achieve that. Only after completing both documents, often a semester-long effort with multiple review cycles, were we...
Deno: What happened? When Ryan Dahl introduced Deno at JSConf EU 2018, it generated significant interest in the JavaScript community. As the creator of Node.js, Dahl presented a compelling vision: a new runtime that would address the common pain points developers face. Deno (a palindrome of "node") promised a fresh approach with new core APIs, ESM-only modules, familiar web APIs, independence from npm, and security by default. Who better to reimagine the JavaScript runtime than the person who...