- Tubelator AI
- >
- Videos
- >
- Gaming
- >
- Perplexity AI 创始人 Aravind Srinivas 与 Stripe 高管 David Singleton 的深度交流
Perplexity AI 创始人 Aravind Srinivas 与 Stripe 高管 David Singleton 的深度交流
在这次炉边谈话中，Perplexity AI 的创始人Aravind Srinivas与Stripe的David Singleton就Perplexity的创立原因、公司现状等话题展开讨论。欢迎加入，了解更多有关AI搜索引擎和企业发展的内容。
Install Tubelator On Chrome
Video Summary & Chapters
No chapters for this video generated yet.
Video Transcript
Hey everyone, thank you so much for joining us and a very warm welcome to our special guest today
Arivan Ditts-Srivenos of
Perplexity AI your CEO
Really excited to have a rich conversation here and I'd first like to learn a bit more about
Perplexity myself and then we'll open up for some Q&A from the audience
So I have been tell us a little bit about the journey. Why did you start perplexity?
It's an AI powered search engine
lots of search engines are there
and what's going on at the company today?
Yeah, thank you.
Thank you all for coming here.
And we started perplexity about one and a half years ago.
Definitely not to build a new search alternative.
We incredibly adoscious.
And I wish I was that adoscious, but that's not the reality.
We started very precisely to focus on one particular problem
of building a great natural language to SQL tool.
We were very motivated and inspired by search engines and Google's story because we are also academics becoming entrepreneurs.
And that was the only example that we could look at.
So that's slowed into how we approach the SQL problem.
We didn't build the SQL solution as a like a coding co-pilot, but rather as a searching over databases sort of a tool.
And one of the tools we built, one of the prototypes we built was actually something relevant to stripe.
Like we looked at how would people do analytics over their stripe data using stripe sigma.
And we built this natural language to stripe sigma tool because it was a some version of presto, right?
And not everybody knows how to write it.
One of our investors, Nat Friedman was actually using it to do some analytics of his own like stripe data.
So all that was very exciting for us, but we were never finding any big, you know, dopamine or traction from real usage.
It was just like a few hundreds of queries a week.
And we decided, okay, nobody is going to give us their data.
If we are like a random startup, nobody knows anything about us.
So we just had to scrape external data and build a cool demo at scale.
And maybe they look at it and then they would give us some data.
And so we did that by scraping all of Twitter.
Like we built this thing called births equal.
We called it births equal because we are not allowed
to use the Twitter name due to trademark.
But it was just literally scraping all of Twitter,
organizing into a bunch of tables
and powering search over that.
And that worked really well.
And that's how we got all of our initial investors.
All that somewhat inspired by how striped
like Patrick and John raised money,
they would show the demo to people
and get like these cool angels, like teeter-teal or Elon Musk.
If you look at Stripes, Angel Investors visits pretty amazing.
So that's how we got a bunch of cool investors, including Jeff Dean.
He tried our Twitter search demo and he was like,
I've never used something like this before and I really like it.
At that time, he did not see anything similar to what we were doing today,
