Rahul Jain
speaker
371 appearances
4 recordings
1 series
first heard Apr 2026
last heard 23 Apr
Rahul Jain’s voice in public audio — every appearance, attributed to the second.
Trend
recordings per month · last 12 monthsRecordings per month over the last 12 months — 4 in all, peaking in Apr 2026 with 4.
Appearances
Like that's a higher priority for us.
And I I think they're part of the reason we ended up pivoting was we also could see it in our own buying behavior where I was reaching out and I was using ChatGPT and and Google's AI overview and AI mode much more than I I was before.
It kind of became the default for me.
And I think that for me is like kind of the trigger when I see myself using a product or a service over and over.
And I'm not typically a a super early adopter.
And I was like, okay, there's clearly something that's happening here in the market.
And you know, we've grown up with search, but this is a completely new experience that I think is really exciting that we're gonna, you know, get to kind of live through.
Yeah, absolutely.
So when you go to Chat GPT, when you go to Google and you effectively just type in that search prompt, what's happening behind the scenes is that these LMs essentially have training data that they're using.
They have essentially snippets that they've accumulated.
And you know, there then there's rap.
And so what they're doing is they're essentially looking at these little chunks of information or data that can best answer that prospect's question.
And so what's happening behind the scenes is that they're taking that data that the LMs have and they're synthesizing it and they're trying to create an answer that answers that that prospect's question to the best of their abilities.
We don't know exactly all the inputs that go into that data set that the LMs are using.
But the hypothesis we had by Noble with now some data behind it was that the the citations that the LMs provide must be somehow related to the the information that's ultimately presented as far as the sources that they're getting this information from.
It's not like a typical like um like a journal where it's like every word is cited and you know where the information comes from.
Typically for a research prompt, there's about 10 to 15 um sources or citations that are provided.
There are many other pieces of information that these LMs are pulling from.
But one thing that we do have, you know, pretty significant confidence in is that the sources are a meaningful input into that answer.
There's definitely a lot of other information that is evaluated that is part of the model.
Showing 41–60 of 371 · page 3 of 19
← Previous
Next →