Databricks, a firm with headquarters in San Francisco and a $38 billion valuation at the time, released open-source technology on Friday for companies to utilize in creating chatbots that resemble OpenAI's ChatGPT.The programme is an artificial intelligence (AI) model, an algorithm that can learn from new data after being trained on sets of data to carry out a variety of tasks.

The publication, according to Databricks CEO Ali Ghodsi, intends to demonstrate a practical alternative to investing a significant amount of time and resources in training a big language artificial intelligence model. A huge language model powers ChatGPT, a well-known chatbot from OpenAI.  An enormous amount of data is used to train OpenAI's $29 billion AI models on a supercomputer owned by investor Microsoft Corp. The costs of computing are "eye-watering," according to Sam Altman, CEO of OpenAI.

Businesses must pay OpenAI to utilize its models for their own applications; the company hopes to earn $1 billion in revenue by 2024.  The work of Databricks comes with warnings. Ghodsi told Reuters that although the open-source chatbot demonstrated impressive abilities when writing blog posts, the company had not yet made official benchmark tests to demonstrate that the bot's performance was comparable to ChatGPT's.

Businesses can purchase cloud-based data mining and analytics software from Databricks, which announced last year that its annualized revenue had surpassed $1 billion.Databricks wants companies to train their own AI models using its platform. Ghodsi asserted that utilizing a tiny amount of data and a computer that anyone with a credit card may rent, the company's researchers educated a two-year-old, publicly accessible model for three hours. According to Ghodsi, everyone will have their own model in the future and will be able to train and enhance it. They would also be spared from having to disclose their data to anyone else.

The decision by Databrick comes as large tech firms like Alphabet's Google and Meta Platforms compete to make AI models smaller, cheaper, and more accurate while also raising millions of dollars in venture capital investment from startups.

Ghodsi projected that these models would ultimately shrink and turn into open-source software. They'll all have them, she said.

Galactic Jellyfish imaged by the Hubble Telescope

Single-play audio messages for Android users are a new WhatsApp feature

(ISRO) launched its LVM3-M3/OneWeb India-2 mission with 36 satellites