The SQL assertion above is a quote from our current Workplace Hours with Benn Stancil. It’s not a SQL assertion that will work at the moment in a cloud information warehouse. However an LLM would perceive it : summarize the guide Moby Dick in two sentences.
Positive sufficient, ChatGPT solutions the query :
This pseudocode blends the structured queries of information evaluation with the unstructured information contained in a basic novel. That is how Benn views the way forward for BI
BI’s Third Kind. “Is the output numbers in the way forward for BI? I don’t assume it’s numbers…the idea is what we would like ; the characterization & qualitative elements.”
Simply as ChatGPT summarized the essence of Moby Dick, AI simply summarizes the themes of 10 buyer conversations – or 100.
Benn views the way forward for BI because the narrative, the story behind the info. Numbers have been the historic supply of reality, a proxy to underlying buyer conduct. Nobel laureate Shiller wrote concerning the energy of narrative economics, tales used to know a enterprise & form the long run.
The remainder of the Fashionable Information Stack will rework, too. Information High quality might want to deal with ever higher volumes. If LLMs do common the perception in new underlying unstructured information sources that are probably huge, what’s Information High quality’s new function on the earth of AI? ETL, extra deterministic in nature, might not see full delegation to AI.
We coated the Databricks/Tabular acquisition & many different subjects. Benn’s views on the info ecosystem are distinctive : a bit irreverent with plenty of sensible expertise underpinning them.
I hope you benefit from the dialog as a lot as I did. The video is embedded beneath & podcast hyperlink is right here.
