The second part of the article discusses web personalization, comparing it to the early hype and subsequent disappointment of blockchain technology. It highlights Google's efforts to personalize search results and suggests that personalization is most useful in shopping, news, and local searches. The author concludes by suggesting that if voice and personalization technologies can overcome current challenges, they could significantly impact the future of technology and marketing.
Key takeaways:
- Super forecasters, who have significantly higher success rates at predictions, change their minds gradually based on new evidence.
- Voice assistant devices are not performing as well as expected, with the majority of use cases being simple commands like playing music or checking the weather.
- Large language models (LLMs) can potentially improve the quality of voice assistants by better understanding user intent and providing more natural responses.
- Web personalization has been a disappointment, with users disliking the feeling of being tracked. However, Google has started to offer more personalization for shopping, which is less controversial and easier to monetize.