Last month I wrote that in order for ChatGPT and Google Bard to reach their full potential, the respective underlying LLMs would need to be trained with industry-specific data from companies such as Bloomberg, IHS/S&P Global, Cerner/Oracle, etc. Those companies ought to be compensated for granting access to ChatGPT and Google Bard (Google has since struck partnerships announced at I/O last week. Read our Google I/O article HERE).

My view is that Bloomberg and other industry-specific data companies would be better served developing their own Natural Language Processing (NLP) front-ends to where users could run text or voice-driven queries against various data sets. Companies such as Zillow, CoStar, Verisk, Solera and scores of other data-centric, Info Services companies would do well to this end. One could use Apache OpenNLP or NLP services made available through APIs from Google Cloud or Microsoft Azure to build the underpinnings of an NLP-based service.

• Asset Management
• Investing
• Software
• Strategy