As the producer of ChatGPT strives to woo more businesses to use its technology, OpenAI aims to roll out significant changes for developers next month to make it easier and faster to construct software applications based on its artificial intelligence models, according to individuals briefed on the plans. The changes include the developer tools’ ability to use memory storage for AI models. This presumably addresses a big problem for partners whose expenses of employing OpenAI’s advanced models might soon pile up as they try to create long-term businesses by creating and selling AI software. It could theoretically reduce costs for application developers by as much as 20 times.

The business also intends to introduce new tools, such as vision capabilities, that will allow programmers to create software that can analyze and describe images, with potential uses in industries ranging from entertainment to medical. Adding these capabilities demonstrates the company’s desire, as Chief Executive Sam Altman envisioned, to become more than just a consumer phenomenon by delivering a popular developer platform. Elon Musk and Altman co-founded the organization in 2015 as a non-profit, where it labored in relative obscurity outside of the tech sector. Currently, Musk has no ownership interest in the business.

On November 6, in San Francisco, OpenAI is scheduled to have its inaugural developer conference, during which the new capabilities are anticipated to be unveiled. According to insiders who requested anonymity to discuss the company’s internal intentions, they are intended to encourage businesses to adopt OpenAI’s technology to create AI-powered chatbots and autonomous agents that can carry out activities without human interaction.

The business made its debut in the market last November with the release of ChatGPT, luring hundreds of millions of users to try out the chatbot that reacted to queries and instructions in human-like ways, making it one of the consumer apps with the quickest rate of growth in the whole world. OpenAI has big expectations for its business. OpenAI executives predicted that this year would end with $200 million in sales and that by 2024, they will have made $1 billion, as originally reported by Reuters in December.

The corporation has had significant difficulties in attracting outsiders to create enterprises utilizing its technologies in recent years. Altman’s top strategic goal is to make OpenAI necessary for other businesses creating apps. He has spoken with programmers to discuss his aim to create a new ecosystem based on the OpenAI models already in many different apps, such as DoorDash and the writing helper Jasper.

Companies will be able to construct apps more affordably by remembering the history of customer conversations thanks to the upcoming release of the so-called stateful API (Application Program Interface). The quantity of use that developers must pay for may be significantly reduced as a result. According to pricing on OpenAI’s website, processing a one-page document using GPT-4 might now cost 10 cents, depending on the length and complexity of the input and output.

A few weeks after the capability was made accessible to ChatGPT users, a further upgrade called vision API would enable developers to create software to analyze photographs. Giving developers access to this tool is also a significant step in OpenAI’s development of what is known as multi-modal capabilities, which process and produce material other than text, such as photos, audio, and video.

MAINTAINING DEVELOPERS’ JOY

These updates are made to persuade additional programmers to pay to use OpenAI’s model to create their own AI applications for various purposes, such as writing assistance or chatbots for customer support.

According to PitchBook statistics, investors have invested more than $20 billion in AI startups this year, many of which rely on OpenAI or the technology of other foundation model companies.

Investors are concerned that these businesses’ reliance on OpenAI and Google might expose them to imitation by rivals or larger corporations through product upgrades. Startups also attempt to diversify their models, testing OpenAI rivals and open-source alternatives like Meta’s (META.O) Llama. Because of this, OpenAI must set itself apart from wealthy competitors like Google (GOOGL.O).

According to these sources, OpenAI has put much effort into keeping developers satisfied. While ChatGPT has had enormous consumer popularity, OpenAI’s efforts to win over other businesses have not been as effective.

The business hurriedly introduced ChatGPT plugins earlier this year, add-on technologies that let programmers construct apps for ChatGPT. OpenAI anticipated that plugins would give them an advantage over competing chatbots like Google’s Bard and become its answer to Apple’s iOS App Store.

Developers of the top 30 or so “popular” plugins have spoken about an early surge of enthusiasm, followed by a sharp decline in interest. According to Lakshya Bakshi, the popular Scholar AI plugin’s developer, there were roughly 7,000 daily users as of late August. There are roughly 180 million active users on ChatGPT each month. Altman has recognized in the open that work is still to be done. Altman acknowledged earlier this year to a gathering of developers in London that plugins had not taken off in the market.