OpenAI introduced new developer tools on Tuesday, including a real-time speech AI tool, to help with AI application development. This move enables Microsoft-backed businesses to compete in the expanding generative AI sector.
Previously, developers had to go through at least three steps: first, transcribing audio, then running the generated-text model to get an answer to the question, and finally, utilizing a separate text-to-speech model.
Businesses that use OpenAI’s services to develop their own AI apps account for a big portion of the company’s income, thus the release of advanced capabilities is an important selling feature.
OpenAI faces rising competition in multi-modal AI research from huge digital companies such as Google’s Alphabet. Despite this, the company anticipates tremendous growth, with sales expected to increase from $3.7 billion in 2024 to $11.6 billion in 2025, while seeking $6.5 billion in investment at a $150 billion value.
OpenAI has released a new fine-tuning tool that allows developers to improve AI models’ replies with visuals and text based on human feedback. This update intends to enhance visual recognition skills for applications such as visual search and autonomous car systems.
The startup has innovatively expanded its toolkit with two cost-effective solutions: a knowledge transfer system that allows smaller models to learn from larger models, and an intelligent “Prompt Caching” feature that reduces development costs by efficiently reusing previously processed AI text segments.
English
French
German
Russian
Spanish