OpenAI Unveils New AI Tools For Developers, Boosting Speech App Creation And Model Fine-Tuning

OpenAI, the startup behind ChatGPT, launched a plethora of new tools to help developers create applications based on its artificial intelligence technology in an attempt to remain competitive in the generative AI race.

The Microsoft-backed business announced that developers would be able to construct AI speech applications with just one set of instructions thanks to a new real-time tool that is available for testing right now.

Previously, the method needed developers to complete at least three steps: first, they had to transcribe the audio; next, they had to use a second text-to-speech model after running the generated-text model to get an answer to the question.

READ: Sam Altman Denies Mira Murati's Exit And Other Departures Are Tied To OpenAI Restructuring Amid For-Profit Transition Plans

OpenAI's revenue is primarily derived from companies that utilise its services to develop their own artificial intelligence applications, therefore the introduction of new features is a major selling factor.

As internet giants like Alphabet, the parent company of Google, incorporate AI models that can analyse text, audio, and video across their enterprises, the competition has also intensified.

According to Reuters last month, OpenAI anticipates that its income in 2024 will increase to $11.6 billion from an expected $3.7 billion. In addition, a $6.5 billion fundraising effort is underway for the corporation, which may lift its valuation to $150 billion.

OpenAI unveiled a model fine-tuning tool on Tuesday, enabling developers to enhance the responses produced by models using text and graphics after training. Human input, which provides the model with instances of both good and terrible replies depending on its responses, can be a part of this fine-tuning process.

READ: Who Is Mira Maruti, Ex-CTO Of OpenAI, Latest High-Profile Exit From AI Firm Amid Wave Of Resignations

Using photos to fine-tune models will offer them stronger image understanding capabilities, enabling applications such as enhanced visual search and improved object detection for autonomous vehicles, OpenAI stated.

The business also introduced "Prompt Caching," which reduces development expenses by half by reusing portions of the text that AI has already analysed. This tool would enable smaller models to learn from larger ones.