The Data Chief | Episode 64

How to Harness the Explosive Growth of ChatGPT for AI Innovation

Amit Prakash

Co-founder and CTO

ThoughtSpot

Current EpisodeEP64: How to Harness the Explosive Growth of ChatGPT for AI Innovation
00:00
00:00

Episode Overview

Historically, computers haven’t been particularly good at determining user intent from real-world language inputs. But, with the rise of ChatGPT and other large language model technologies, AI systems are achieving new levels of accuracy in understanding and responding to human queries. In this episode, Amit Prakash, Co-founder and CTO of ThoughtSpot shares his thoughts on the evolution of AI, and the responsibility needed to make it ready for everyday use.

Key Moments: 

  • What was the problem Amit was trying to solve when he co-founded Thoughtspot? (04:48)
  • When did Amit first start looking at ChatGPT and natural language processing? (08:16)
  • Why did ChatGPT explode in 2023? (21:00)
  • Is ChatGPT ready for important work and research? (31:26)
  • Will ChatGPT spell the end of platforms like Thoughtspot? (32:41)

 

Key Quotes: 

It's not just words it's a very specific intent behind the question, and can you match that intent with the intent behind the document that you're going to bring up? And both of these things are deeply related to language understanding. And typically computers haven't been really good at understanding language and the real intent behind it.

So the way I think of ChatGPT or that family of large language models is that it's kind of like we've got a new element on the periodic table. And yes, that could be a radioactive substance and it could explode in your hands if you don't handle it properly, but it absolutely opens up the opportunity to experiment and combine it with other elements and build products that were just not possible before.

In particular, one of the things that we've realized working on this for so long is that number one trust is so important in the data space. You cannot put a product in front of people that's supposed to answer data questions and it gets it wrong.

Mentions:

 

Bio:

Amit is Co-founder and CTO at ThoughtSpot and has deep experience in building large scale analytics systems. He holds at least 10 patents related to search and NLP. Prior to ThoughtSpot, Amit led multiple analytics engineering teams in the Google AdSense business was a founding engineer in the Bing team at Microsoft, where he implemented the pagerank algorithms for search from scratch. Amit received his PhD in Computer Engineering from the University of Texas at Austin and a Bachelor of Technology in Electrical Engineering from the Indian Institute of Technology, Kanpur.