Google twilight? OpenAI wants to introduce a new era of research

At this point, there is no confirmation yet whether the OpenAI search engine will be introduced as a new product and under its own name, or whether it will be a ChatGPT variant. At the moment, it is unofficially said that it will be a more expensive version of ChatGPT, which will simply get a real-time network search function and a changed user interface.

Is it possible to compete with Google in a sector that clearly belongs to it?

What's up against OpenAI?

The search engine market is a tough business, including: It's a chicken-and-egg dilemma. The egg is the content or data available online that users want to find, and the chicken is the search engine itself and the opportunities it offers to website creators and owners. And now like this:

• Website creators want their content to be included in a particular search engine if it benefits them in the form of traffic – a lot of users visit the search engine, so there is a chance they will find their content and visit the site. You can then monetize this traffic on your website, for example, through advertising or product sales
• Users want to use a particular search engine if it provides high-quality results, that is, high-quality content for almost any query

Users will visit a search engine if it provides high-quality search results, filters millions of pages and selects the best ones according to complex algorithms. Website builders want to create content that is optimized for search engines with a large number of users – otherwise optimization activities will not be profitable.. If there are no users, website developers are not interested in working with a particular search engine, and if there are no quality websites, users will not be willing to use a particular search engine.

In addition, there is also an assessment of the quality of individual pages. To attract users, a search engine must be able to provide them with access to broad, high-quality content. In contrast, to access a wide range of content, A search engine needs a large group of users who contribute to its development by generating data and signals (such as clicks and queries) that can be used to improve search algorithms and the accuracy of results.. In other words, a search engine needs users to become better, but to attract users, it needs to be really good.

Google occupies the first position in the search engine industry, and the company has worked for many years to reach this position. Depending on the source, Google search engine worldwide Up to 91.54 percent participate I 5.9 million searches are performed every minute. Users use Google and searching for information on the Internet is often associated with the phrase “just Google it.”

Website builders also value Google because the search engine often provides the most traffic and is the dominant source of user acquisition for stores, news sites, and blogs. It's often said that “if your business isn't on Google, you don't exist.” When it comes to where websites get traffic, Google accounts for an average of 66%.

OpenAI lures with money

OpenAI has access to the technology behind Microsoft's search engine: Bing. It's not used as often as Google, but the way it works is quite similar.

Furthermore, to ensure the search engine works, OpenAI must constantly monitor and index the Internet. If something new appears on the web, OpenAI must process it and then decide whether it's worth including the page in its index, and if so, where. The so-called network monitoring is used. Crawlers The company already has its own crawler called GTTBot.

If ChatGPT is to answer a question and base the answer on what it finds on the web, two tasks must be accomplished:

1. The OpenAI crawler should visit the page in question and attempt to connect to and process its content
2. The site must allow this

Here another challenge arises for OpenAI. Website developers and owners are already used to the way things are now, and users are used to using Google to accomplish specific tasks. In the OpenAI search engine, individual website owners must receive appropriate incentives. In Google's case, the incentive is simple: a large number of redirects, i.e. traffic to the website or store. For a search engine other than Google, that incentive is often just money.

OpenAI is already negotiating, among other things: with publishers of large news sites to make their content available to the GTTBot crawler For appropriate fees. Depending on the size of a particular website, these fees can reach millions of dollars. Many large websites are waiting for this kind of “invitation” to collaborate with OpenAI, and here is why Intentionally blocks access to the GTBot crawler on its websites.

Publishers and other website owners block GTBot because they don't like giving their content to AI. The most common reasons are ethical and copyright issues.

If OpenAI convinces enough publishers and large websites to make their content available to the GTTBot crawler, it could be possible to create a search engine that delivers high-quality results. However, it is difficult to say whether it will be profitable, because paying for access to individual content could be very expensive in the long run. However, with this action, OpenAI may cause Google to worry about its dominant position. But are you sure?

Google is betting on artificial intelligence

We already know that despite the introduction of artificial intelligence in the Bing search engine, Microsoft failed to attract many new users permanently and after a temporary increase in traffic, Internet users generally returned to their old habit of “Googling”.

However, Google reacts to the actions of its competitors. The matter was not limited to providing a so-called solution Gemini, a competitor to ChatGPT, also announced the introduction of SGE, i.e. Generative Search Experience.

It's an innovative approach to internet searches It uses advanced AI and language models to create more contextual and insightful answers to user queries.

SGE aims to make it easier for users to find answers to complex questions that may not be directly supported by a single website by providing a so-called “snapshot” that aggregates information from multiple sources and summarizes it to help users understand key aspects of a topic. Google wants to handle more complex queries this way, often consisting not of a keyword phrase of one, two, or three words, but of complete sentences. Currently, SGE is still in the development phase and is currently only operational in the US for a closed group of users as part of the Google Labs project.

SGE is a key priority for Google to advance search using generative AI, which it already has Allowing a more natural and intuitive experience for users. Right now, there's no release date for SGE (it's said to be later this year), but if SGE can efficiently answer complex user questions and engage in conversations with them, then The OpenAI project may have had a very rocky start. This does not mean that it will lose its chance at commercial success even before it is available.

Author: Grzegorz Kobra, journalist at Business Insider Polska