The Future of (Generative AI) Search

Explore the future of search with generative AI. Discover how Apple Intelligence, context-based understanding, intent-driven interactions, and integrated workflows are transforming search. Learn about the trust challenges and the critical balance needed for reliable, AI-powered search experiences.

Abstract image of a telesco

In this webinar from June 11, 2024, we dive into the fascinating world of generative search and explore how it's evolving far beyond the traditional notion of search. We kick off by discussing the recent announcements from Apple's WWDC keynote, particularly focusing on Apple Intelligence and how it sets the stage for the future of search.

We then explore what search truly encompasses, going beyond just web searches to include commerce, file systems, and more. By examining different dimensions of search like perception, reasoning, and action, we paint a picture of how AI is transforming search as we know it.

Throughout the webinar, we take a deep dive into three key trends: the progression from language-based to context-based perception, from command-driven to intent-driven reasoning, and from simple information retrieval to integrated workflows and actions. We discuss how these shifts, enabled by advancements in AI, are reshaping the search landscape.

However, we also tackle the critical issue of trust in generative search systems. We examine various trust problems such as contradictory sources, hallucinations, variability across tools, prompt formulation bias, and generative pollution. By presenting a trust matrix, we categorize different search tools and highlight the challenges in establishing trust.

Key Points:

  1. Apple Intelligence marks the start of a new AI-powered era for Apple, with deep integration into the operating system and a privacy-first approach.
  2. Search extends far beyond web searches, encompassing diverse domains like commerce, file systems, and more.
  3. The complexity of perception, reasoning, and action in search systems is evolving, leading to more advanced AI-powered search experiences.
  4. Search is progressing from language-based to context-based understanding, leveraging multimodal inputs and even exploring the potential of neurotechnology.
  5. The shift from command-driven to intent-driven interactions enables search systems to better understand and fulfill user needs without explicit instructions.
  6. Search results are moving from simple information retrieval to integrated workflows, seamlessly connecting information across apps and systems.
  7. Trust is a critical challenge in generative search, with issues like contradictory sources, hallucinations, tool variability, prompt bias, and generative pollution impacting reliability.
  8. The trust matrix categorizes search tools based on the level of trust in their sources and their ability to summarize or synthesize information.
  9. Case studies in domains like law highlight the significant challenges in establishing trust in generative search systems, despite their potential for efficiency gains.
  10. The future of search lies in striking a balance between leveraging the power of generative AI while addressing the fundamental issues of trust and reliability.

Following are the slides and an edited transcript of our research webinar from June 11, 2024.

Transcript: Hello, all. Thanks for joining us. We're excited to have you here. We're going to be talking a bit about generative search - search in general - which we're going to walk you through why we see it as much more than what you might think of just as search. And we will kick things off by covering and talking a little bit about what happened yesterday at the keynote of Apple's WWDC conference, as it puts a lot of the search stuff in perspective.

Transcript: What was announced yesterday shouldn't be a complete surprise to you. If you've been following this for a little while, we've talked a lot about the dynamic of moving AI to the edge to be able to run on device, which was a key part of the presentation yesterday. We've talked a lot about how AI will be pervasive, will be always on, will know you intimately as it gets to understand a lot more about you through all of the interactions on your phone or device. We've talked about how Apple might abstract some of the user flow away from Google by answering some of those questions in chats, etc. So this felt like a good day to see these announcements and think that we interpreted a lot of the product moves and the research and science that Apple has published pretty well.

Transcript: Some key things to talk about in terms of what we see that came out of the presentation yesterday - we're just going to focus in on the Apple Intelligence stuff. There were a lot of other interesting things that came out, but the interesting part is this Apple Intelligence is in a lot of ways sort of the start of a new aiOS for Apple. They are embedding this technology throughout the operating system, and making it available through the developer frameworks and toolkits and SDKs. There are a lot of new brand names of things that are related to these technologies, and availability for developers to embed this inside of their applications, which is interesting.

Great! You’ve successfully signed up.

Welcome back! You've successfully signed in.

You've successfully subscribed to Artificiality.

Success! Check your email for magic link to sign-in.

Success! Your billing info has been updated.

Your billing was not updated.