Google plans to launch a new chatbot search later this year. How will this work?
Google’s biggest challenge with a chatbot search is transfer of responsibility: At this time, Google provides links to answers provided by website publishers. With a chatbot search, it would provide direct and authoritative answers to questions.
Today, Google is already partly responsible for the answers to search queries: by prioritizing search results, Google influences the search for information. With No-Click Search, Google confidently cites more or less appropriate snippets of web pages for search queries.
But it will be with chatbot research, slated for 2023 according to The New York Times, that Google will have to take full responsibility for every answer it provides — at least in theory. More on that later.
The pitfalls of chatbot research
Beyond monetization, there are at least four key challenges for Google:
- The chatbot information cannot be wrong. Otherwise, the chatbot is not trustworthy for a search query, and therefore redundant or even harmful.
- Even if Google were able to create a reliable search chatbot, taking responsibility for the answers (“Google said that…”) would be a major challenge at the scale at which Google operates. Even correct answers could be misinterpreted.
- google needs a system that integrates site operators and content creators. Over time, they provide the information the chatbot needs to generate responses. They are also Google’s advertising customers through search and display ads.
- As Google integrates site owners and content creators into chatbot search, it needs to do so in a way that preserves copyrights and allows publishers to enjoy. The years-long debate on an EU-wide ancillary copyright for press publishers shows how difficult this task is. With a chatbot, it could get even bigger.
Google could, like OpenAI with ChatGPT, launch an experimental, standalone chatbot that offers many other features besides answering questions, is not explicitly designed as a search engine, and therefore does not have to fully answer all above requirements.
This chatbot could be Deepmind’s Sparrow, cutting the veils of Microsoft and OpenAI and reassuring shareholders. It would be a short-term solution.
Google will optimize clickless search with AI
In the long term, I think it’s more likely that Google will use AI to gradually optimize existing search, always expecting higher levels of monetization. Search has been Google’s growth engine and cash cow for years, generating billions in revenue quarter after quarter. Google is not going to take a financial risk and reset its search engine.
Instead, Google will first identify areas where it has a high likelihood of providing competent answers. Language models such as Med-PaLM, which are tailored to specific catalogs of questions, can answer reliably at the level of human experts.
Google could train such models for different categories and refine them through user feedback. In categories that AI can reliably serve, Google will experiment with new ad formatssuch as ad links directly in AI-generated responses.
Additionally, Google will use AI methods to reliably summarize multiple sources. He will incorporate these summaries and quotes into his responses to the AI, trying to provide a human source alongside each AI claim. This way, Google could transfer at least some of the blame to the cited sources, but would still benefit from bringing more attention to its platform.
At the same time, Google would become less dependent on the content that website operators and publishers provide for search. Google’s AI responses would also likely increase the number of interactions and time spent on its platform, increasing its advertising value.
Meta’s scientific AI model, Galactica, has shown that there are still technical issues to be resolved with citations. However, Google may set stricter guidelines.
Perplexity.ai gives you a glimpse into the future of Google Search
If you’re wondering how such a combined chatbot-quote search might work in practice, head over to the Perplexity.ai website: the experimental chat search engine generates a question answer based on the content of the site’s operators Web and name other sources.
For example, if I ask about the best VR headset in 2022, it tells me Meta Quest 2, gives a reason for its recommendation, and points to other sources. It works well on the surface, but falls apart as soon as I ask a deeper question: Should I really buy the Quest 2?
Quest 2 is the best VR headset, Plerplexity.ai replies, but this one advises against buying it because it’s made by Facebook – and it refers to the opinion of a single tech editor.
In addition, Perplexity’s recommendation changes with each new request, even if the content remains similar. The experimental search engine is therefore only suitable as an interface demo, a proof of concept. That’s how it could work. And Google? He could do better.
With a step-by-step approach, using specialized models and immense data training, combined with user feedback data that only Google has at this scale, the research company could Gradually resolve content issues for more and more categories and respond more and more reliably with AI.
This process will be smooth, it could take years and publishers will suffer. I expect big discussions on copyright and regulation.