OpenAI CEO Sam Altman and other key executives recently held a Reddit AMA to discuss ChatGPT Search, offering insights into its functionality and potential impact on SEO. This panel included:
ChatGPT Search is distinct from a conventional search engine; rather, it’s an AI chatbot with search capabilities. While it doesn’t compete directly with Google as a search engine, it offers a fresh way for people to find information, making it useful for work and daily life.
Importantly, ChatGPT Search doesn’t show ads or track users, building on the trust ChatGPT has already earned with users for both personal and professional use.
One reason ChatGPT Search may be perceived as a challenge to Google is that users are already familiar with ChatGPT and have positive associations with it, which makes it easier for them to integrate it into their routines without needing to “switch” from Google.
Sam Altman on ChatGPT Search’s Advantages
During the AMA, a Redditor asked Altman what makes ChatGPT Search valuable compared to other search engines. Altman responded:
“For many queries, I find it to be a way faster/easier way to get the information I’m looking for.
I think we’ll see this especially for queries that require more complex research. I also look forward to a future where a search query can dynamically render a custom web page in response!”
This comment about a “custom web page” hints at future personalization that could deliver results uniquely tailored to each search query.
Handling Complex Queries
Altman highlighted ChatGPT Search’s advantage with complex queries, as it relies on natural language input.
Unlike Google, which often requires keyword-specific queries, ChatGPT can interpret conversational language, making it more intuitive for users.
Grounding Information for Improved Accuracy
The AMA also touched on OpenAI’s approach to handling hallucinations (when ChatGPT generates inaccurate responses) and incorporating up-to-date information.
This is achieved with a technique called Retrieval-Augmented Generation (RAG), which selects data from updated sources like a search index or knowledge graph, providing the model with reliable information to form responses.
One Redditor asked about whether hallucinations would persist and how outdated information would be managed. Mark Chen, OpenAI’s SVP of Research, explained:
“We’re putting a lot of focus on decreasing hallucinations, but it’s a fundamentally hard problem – our models learn from human-written text, and humans sometimes confidently declare things they aren’t sure about.”
Chen mentioned that Reinforcement Learning (RL) might help reduce hallucinations by reinforcing factual responses.
In RL, the model is “rewarded” when its answer is correct, reinforcing the behavior over time. Chen noted:
“Our models are improving at citing, which grounds their answers in trusted sources, and we also believe that RL will help with hallucinations as well – when we can programmatically check whether models hallucinate, we can reward it for not doing so.”
ChatGPT Search and Bing
A Redditor asked if ChatGPT Search still uses Bing as its search engine. Srinivas Narayanan, OpenAI’s VP of Engineering, confirmed:
“We use a set of services and Bing is an important one.”
This suggests that while Bing plays a key role, ChatGPT Search uses multiple sources to retrieve data, though exactly which sources remain unspecified.
SEO for ChatGPT Search
When asked about SEO optimization for ChatGPT Search, Chief Product Officer Kevin Weil acknowledged the novelty of the product:
“This is a great question—the product just launched today so there’s a lot to figure out still about where search will be similar and where it will be different in an AI world. Would love any feedback you have!”
Key Takeaways for SEO and ChatGPT Search
1. Bing is a key data source for ChatGPT Search, but other services are also used, making Bing an important search engine to consider for rankings.
2. Natural language queries are an advantage, as users familiar with ChatGPT’s conversational style may increasingly use ChatGPT Search.
3. Reinforcement Learning may reduce hallucinations, improving answer reliability over time.
4. Personalization might be in the pipeline, potentially with dynamically rendered pages tailored to user needs.
OpenAI’s new approach isn’t a direct competitor to Google but offers a unique and evolving experience for finding information on the web.