How can OpenAI realistically achieve such a level of usefulness and ubiquity?

[Dimaeiya333]

If OpenAI is to compete with Google, it will have to offer a highly useful product that will attract users to use it regularly. For example, Apple and Google have a committed audience in the mobile ecosystem that regularly incorporates them into their daily lives. Both at work and at home. It is absolutely clear that the competition is not enough to simply create a new search engine .


OpenAI is currently facing an uphill battle against Google, but also Apple and spain mobile database Microsoft. If we count apps and other internet devices, add Amazon, and include other competitors that are integrated into the daily lives of billions of users, it has a lot to do.

OpenAI doesn't have the power to launch a search engine that would compete with Google. It doesn't have an ecosystem that would support integration into users' lives.

OpenAI lacks expertise in information retrieval
Searching research papers and patents is formally referred to as information retrieval (IR). No search of the research paper repository at Arxiv.org will turn up papers authored by OpenAI researchers related to information retrieval.

The same can be said for patent searches related to information retrieval (IR). OpenAI’s research paper list also lacks studies related to information retrieval.

This certainly doesn't mean that OpenAI is secretive. OpenAI has a long history of publishing research papers on the technologies they develop. But IR research doesn't exist. So if OpenAI wants to compete with Google, where do we see the smoke rising from this fire?

There is currently no sign that OpenAI is developing any search, or even flirting with building a new search engine. There is clearly no such thing.

Is the OpenAI search engine a Microsoft project?
There is evidence that Microsoft is investigating how to use LLM as part of a search engine. All other research work is classified as belonging to the field of artificial intelligence, natural language computing, and information retrieval (aka search).

Here are some research papers from 2024:
Extracting structured entities using large language models:
This research paper investigates a way to extract structured information from unstructured text (such as web pages).

In practice, this means converting a web page (which represents unstructured data) into a machine-understandable format (which is structured data).

Improving human annotation: Leveraging large language models and efficient batch processing:
This is the use of artificial intelligence for search query classifications.

Improving text input using large language models:
This research paper investigates how to obtain high-quality text embedding that can be used for information retrieval (IR). Text embedding refers to creating a representation of text in a way that algorithms can use to understand the semantic meanings and relationships between words.