Will SearchGPT change the way we search?

Hi!

Search engines were a game changer when they were introduced in the 1990s. No more leafing through encyclopedias to check facts or visiting the local library to research topics. 

Answers to questions could be found quickly and easily online without needing to speak to experts or professionals. And on top of this, people were able to discover new websites and content.

Internet use changed considerably as a result. You no longer needed to know the URL of a particular website in order to access it or use web directories to find new sites and navigate the internet. 

Search engines opened up a world of information and have transformed the way we access and consume information to the point that ‘google’ has even become a verb.

With the advent of AI are we about to see another shift in the way we use and access information?

Enter SearchGPT

OpenAI has been working on AI search features and their prototype, SearchGPT, is currently being tested by a limited group of users and publishers. 

The aim is to incorporate some of these search capabilities into ChatGPT to not only provide fast answers based on up-to-date information but also gather and provide relevant links to the source material.

OpenAI has also made a commitment to working with publishers and creators to ensure proper credit and use of their content - they already have partnerships with Associated Press and The Atlantic among others. 

With concerns over AI’s impact on copyright and intellectual property, these partnerships may be a small step towards protecting content creators and alleviating the concerns of writers and publishers.

The race for AI search dominance

It’s not the first AI-powered search engine though. Microsoft released an AI-powered version of Bing in 2023 - now known as Copilot - which made use of OpenAI’s LLMs. It hasn’t exactly had a roaring success but shows clear intentions to further integrate AI in search.

It will be interesting to see how the king of search engines, Google, responds to the threat from SearchGPT. Google’s attempts to incorporate AI into searches have been a little hit and miss so far - they were subject to a bit of ridicule when their AI Overview feature suggested users add non-toxic glue to make cheese stick to pizza.

Despite the odd misstep they are pushing ahead with their AI development. As the dominant search engine with over 90% of search engine market share worldwide they certainly won’t want to be relegated to the sidelines.

The pressure to stay ahead, or at least not get left behind, means companies are putting large amounts of time and money into developing their own versions but rushing to market without proper testing and refinement has consequences. 

Innovation needs to be balanced with responsible development and ethical implications. The true measure of success will lie in user adoption and trust, not just technological advancement.

New players in the game 

It’s not just the existing big boys fighting it out, new companies are also throwing their hats in the ring. Perplexity already provides a concise, summarized answer to search queries with links to source material if users want to delve further to check facts or learn more.

The appeal of AI search

Part of the appeal of tools like Perplexity and AI chatbots like ChatGPT and Gemini is the speed of response. Rather than trawling through pages of search results yourself, an AI assistant can summarize the information and even rephrase or simplify it to help you with comprehension. If you have follow-up questions they can go into further detail in specific areas if needed.

There’s also added potential for contextual understanding. Rather than relying on keywords and algorithms to determine results, AI can be used to interpret the intent behind queries which should result in more relevant results for users. 

Challenges and limitations

There are drawbacks to current AI-powered search options though. If you were to use an AI model in the same way you currently use a search engine, you might find that the information you’re getting isn’t up-to-date as the model has only been given training data up to a certain point in time. There’s also a risk of errors and inaccuracies in the results as well as potential for bias. 

Combining real-time information and making improvements to mitigate errors and combat bias may be the key to success. Nailing this aspect will likely play a large part in determining which AI-powered search tools become dominant in the future. It will be interesting to see how SearchGPT handles this challenge!

In the meantime, as AI-powered search tools continue to evolve, it's worth keeping an eye on these developments - they may just revolutionize the way we find and process information online.

If you’re interested in giving SearchGPT a go, you can join the waitlist here. 

OK, speak soon!

Inge
Top Apps Founder

P.S. Check out the Top Apps AI courses here.