OpenAI Unveils New AI Development Tools

(Reuters) – OpenAI unveiled a host of new tools on Tuesday that would make it easier for developers to build applications based on its artificial intelligence technology, as the ChatGPT maker wrestles with tech giants to keep up in the generative AI race.

The Microsoft-backed startup said a new real-time tool, rolling out immediately for testing, would allow developers to create AI voice applications using a single set of instructions.

Previously, developers had to follow at least three steps: transcribing audio, generating a response using a text model, and then applying a text-to-speech model.

A significant portion of OpenAI’s revenue comes from businesses utilizing its services to create their own AI applications, making the rollout of advanced capabilities crucial for sales.

Competition has intensified as technology giants like Google-parent Alphabet integrate AI models that can process various types of data including video, audio, and text across their platforms.

OpenAI projects its revenue to soar to $11.6 billion next year, up from an estimated $3.7 billion in 2024. The company is also undergoing a $6.5 billion fundraising effort that may value it at $150 billion.

As part of Tuesday’s launch, OpenAI introduced a tool for fine-tuning models after training to enhance responses generated from images and text. This process allows developers to use human feedback to showcase examples of strong and weak responses.

Using images to fine-tune models improves their understanding, enabling advancements in applications like enhanced visual search and better object detection for autonomous vehicles, according to OpenAI.

The startup also revealed a tool for smaller models to learn from larger ones and “Prompt Caching,” which reduces development costs by half by reusing previously processed text.