Databricks is buying generative artificial intelligence (AI) startup MosaicML in a $1.3 billion deal.
The acquisition is designed to meet the rapid demand among businesses for their own version of ChatGPT, The Wall Street Journal (WSJ) reported Monday (June 26).
Officials for Databricks — a data storage and management firm — said in the report the deal melds its AI-ready data management technology with MosaicML’s language model tool.
Companies can create their own low-cost language models using proprietary data, rather than the current system, in which most businesses use third-party language models trained on widely-available online data, per the report.
“If you’re building a model from scratch, you know what you’re feeding it,” Databricks CEO Ali Ghodsi told the WSJ.
Using third-party models, he added, can give companies extraneous information that can skew results and expose them to privacy and security concerns due to the sharing of data in models built by external vendors.
PYMNTS looked at the way the usefulness of AI is often overshadowed by the hysteria around the technology in a conversation with Jeremiah Lotz, managing vice president of digital and data at PSCU.
Lotz told PYMNTS he believes that generative AI offers exciting potential for use in helping generate more personalized and natural interactions between consumers and financial services providers.
Within the payments space, generative AI can have an effect by improving speed, security, convenience and personalization via real-time analyses of behavioral preferences around the payments occasion.
However, the potential uses of generative AI don’t end there. Lotz argued that combining generative AI with voice and speech capabilities would usher in a new world of interactive, intelligent and conversational commerce.
“Generative AI, whether over voice or text, will give a more unique conversation and get into more of a dialogue of not just, ‘Do you offer a savings account?’ but allowing users to say, ‘I’m looking to save money for my 8-year-old for college,’ and have the bot respond with the right types of savings or investment vehicles, not just a static list,” he said.