OpenAI Reveal SearchGPT as Direct Challenge to Google and Perplexity

On Thursday, Open AI unveiled its much-anticipated search engine prototype, SearchGPT, which aims to give users “fast and timely answers with clear and relevant sources” from web sources. According to OpenAI, they took inspiration from ChatGPT’s UI and built SearchGPT around a standard search box that takes natural language queries as input.

We’re testing SearchGPT, a temporary prototype of new AI search features that give you fast and timely answers with clear and relevant sources.

We’re launching with a small group of users for feedback and plan to integrate the experience into ChatGPT. https://t.co/dRRnxXVlGh pic.twitter.com/iQpADXmllH

— OpenAI (@OpenAI) July 25, 2024

For example, when you insert a prompt, SearchGPT will select information and photos from the web along with links to relevant sources and generate AI-generated text to answer the user’s question directly. To not discredit the source, OpenAI mentioned that its new search engine will cite the webpage on which a prompt response is based with a link placed in parentheses at the end of the text.

In addition, the company mentioned that SearchGPT’s prototype is to collect feedback from a group of initial adopters and then integrate the best features into ChatGPT. Only 10,000 test users can access the search engine at launch as they are working with third-party partners or using direct content feeds to build search results. 

Related: ChatGPT-5: Release Date, Subscription Price and Everything Else We Know So Far

How Does SeachGPT Work? 

As mentioned above, SearchGPT follows most of ChatGPT’s UI and its models, specifically GPT-3.5, GPT-4 and GPT-4o. Thus, when you use the search engine, you will see a large textbox asking, “What are you looking for?” Then, after typing in your query, SearchGPT serves up information and photos from the web along with links to relevant sources. Moreover, you can ask follow-up questions or explore additional, related searches in a sidebar.

For instance, OpenAI showcased that the search engine summarises its findings on music festivals and then presents short descriptions of the events followed by an attribution link. According to OpenAI, the service doesn’t simply answer the additional queries entered by the user but also accounts for the earlier requests that preceded them. That context can help the underlying AI models generate more relevant output. 

Moreover, based on The Verge, there is also a feature called visual answers, and some searches may ask for your location as SearchGPT “collects and shares” general location information with third-party search providers to improve the accuracy of results. 

In their blog post, the company wrote: “Getting answers on the web can take a lot of effort, often requiring multiple attempts to get relevant results. We believe that by enhancing the conversational capabilities of our models with real-time information from the web, finding what you’re looking for can be faster and easier.”

Is SearchGPT the Google Killer?

OpenAI has long since hinted at ideas of overthrowing Google, with the most recent report coming in February when The Information stated a web-search product was in development. While there is no confirmation if it is SearchGPT, the release of the search engine prototype is not the most optimal timing as AI-powered search tools are under justifiable fire for plagiarism, inaccuracies and content cannibalism.

Google stock immediately plummeting after the @OpenAI SearchGPT demo 😬

Anyone who has been paying attention knows there will be a new king of search within 10 years.pic.twitter.com/asnjzSaLxi pic.twitter.com/GCe3l9lmgi

— Alex Valaitis (@alex_valaitis) July 25, 2024

For instance, Google’s AI Overviews infamously recommended people put glue in pizza. In addition, Perplexity, an AI answer engine startup, came under criticism for its AI summaries feature that publishers claimed was directly ripping off news articles written by other outlets, including CNBC, Bloomberg, and Forbes, without giving credit or citation. Not to forget, OpenAI also had a law dispute with the New York Times over alleged unethical practices in March.

However, despite the rough timing of the release and previous accusations, OpenAI has firmly stated that SearchGPT “is to help users connect with publishers by prominently citing and linking to them in searches. Responses have clear, in-line, named attribution and links so users know where information comes from and can quickly engage with even more results in a sidebar with source links.”

OpenAI has also mentioned that publishers will have a way to “manage how they appear in OpenAI search features”, meaning they can opt out of having their content used to train OpenAI’s models and still be surfaced in search.

Meta Is Also Not Sitting Back with Its New Llama 3.1 

Besides OpenAI unveiling SearchGPT, Meta has unveiled Llama 3.1 405B, the first frontier-level open-source AI model. In a blog post by CEO Mark Zuckerberg, he stated that several tech companies are developing leading closed models. However, open-source is quickly closing the gap, and Meta’s new Llama 3.1 will become the most advanced in the industry by leading on openness, modifiability, and cost efficiency.

Starting today, open source is leading the way. Introducing Llama 3.1: Our most capable models yet.

Today we’re releasing a collection of new Llama 3.1 models including our long awaited 405B. These models deliver improved reasoning capabilities, a larger 128K token context… pic.twitter.com/1iKpBJuReD

— AI at Meta (@AIatMeta) July 23, 2024

For instance, the upgraded 8B and 70B models will feature multilingual languages, a significantly longer context length of 128K, state-of-the-art tool use, and better reasoning capabilities to form long-form text summarisation, multilingual conversational agents, and coding assistants. Attached is a picture from Meta’s blog, comparing its LLama capabilities with ChatGPT.

In addition, to stay true to their open-source claim, other developers can utilise the 405B and more to improve their own models, such as real-time and batch inference, supervised fine-tuning, evaluation of your model for your specific application, continual pre-training, retrieval-augmented Generation (RAG), function calling and synthetic data generation.

As to whether an open-sourced AI model will pave the future, it remains to be seen, but Zuckerberg is confident as he ends off with: “I believe the Llama 3.1 release will be an inflexion point in the industry where most developers begin to use open source primarily, and I expect that approach to only grow from here. I hope you’ll join us on this journey to bring the benefits of AI to everyone in the world.”

Also Read: Meta Quest 4 and Quest Pro 2 Hint at 2026 and 2027 Release Window 

The Race for Search Engine Optimisation Ownership Continues

In our modern and digital age, the importance of search engine optimisation (SEO) has never been greater. As businesses and content creators strive to gain visibility and attract audiences online, the race to optimise for search engines like Google, Bing, and now OpenAI is intensifying.

However, the race for SEO development is a marathon, not a sprint. As businesses and content creators strive to enhance their online presence, the future of SEO will rely on greater integration of AI and machine learning for a more sophisticated understanding of user intent and further emphasis on user experience.

For more trending tech and AI-related news, follow our Twitter for timely updates and more recent technological advancements trying to improve our lives.