Jain: Language models were originally built specifically to improve Google Search
Arvind Jain · He quit his cozy Google job & ignored lean startup advice— then grew to $3M in 1 year. | Arvind J... · PMF Show · Apr 29, 2024 · at 23:25
Arvind Jain, former Google engineer and CEO of Glean, discusses the origin and initial motivation behind Google's language model research.
0:00 / 0:09exact quote · 9.4s
▶ Watch the full episode on YouTube → 720p mp4 · rendered on demand · StarZero watermark
“All of this research happened in Google. And LLMs were actually, or these language models were actually built, you know, with one purpose, which is to make search better, to make Google search better.”
quote is from the automated transcript, cleaned for reading: filler sounds and stutters are removed, nothing is rephrased. names can be misheard (the analysis reads context, assessments check outside sources). how →
More from Arvind Jain
Jain: Glean intentionally did not charge customers for its first two years
“We actually were not selling the product. We had no pricing. We had no plans to make revenue. That was an intentional decision we made that we were actually not even charged for our product for the first two years.”
Jain: Big Tech safety nets and zero consequences for failure breed complacency
“Cause sometimes I feel like, you know, at Google, I was not feeling challenged anymore. Cause you know, we're such a successful company and we could always, you know, take big problems, you know, and work on them. And it didn't matter if they failed. Cause lik…”
Jain: He avoided early feedback from business people to protect his idea
“In fact, the more people I would talk to, like, you know, especially people who are, Who have business acumen, like they would actually turn, turn me farther from this idea. So I actually chose not to talk to too many people.”
Jain: Glean skipped TAM analysis entirely because the market did not exist
“Like, you know, sometimes we sort of talk to our team about it, so we didn't do any TAM analysis. Like, how do you do TAM? There's no product in the market. Nobody actually is even solving this problem.”
Jain: Avoid using friends as early design partners; prioritize cold outreach instead
“I didn't actually go and, like, you know, hit my friends, you know, and say, like, hey, can you actually, you know, take this product and give me feedback? Because I felt like, you know, they'll all be very nice to me. And they will do it like, you know, even …”
Jain: First-time founders optimize for entrepreneurial success; repeat founders focus on problems
“For a first time founder, like, you know, they're actually trying to, their goal is actually You know, succeed as an entrepreneur. They want to actually learn that, you know, that art. Sort of like, you know, it's personal in that sense. You know, for me, that…”