Who is on this subjectordered by the date of their first quote here
Every lane here carries 5 quotes or more. Offsets are days from the middle first-quote date, 26 Jan 2023 — a date, and nothing else. It is not a claim about who reached a view first.
-
Bill Gurley
10
26 Jan 2023 – 4 Aug 2026
+0 d
- Dylan Patel 10 5 Dec 2023 – 6 Jul 2026 +313 d
The chronologysourced and dated, oldest first
-
-
Bill Gurley
Gurley calculated acquiring tech industry contacts at 70 cents per name from a conference PalmPilot and spammed them.
“People weren't really doing, like, cost of customer acquisition back then, but I think it was, like, 70¢ a name or something like that of the most influential people in the entire tech industry.”
-
Dylan Patel
Patel identifies distributed training across data centers with lower bandwidth as a key unsolved problem that would unlock massive scaling.
“Everything that we've seen so far is that large scale training has to happen in an individual data center with very high speed networking.”
-
-
-
Dylan Patel
Patel says most people will respond to outreach but almost nobody tries.
“One of things I learned early on is you can talk to anyone on the Internet. You can just reach out to anyone, and most people just say yes, but no one does it.”
-
Dylan Patel
Patel advises becoming a foremost expert in what you enjoy and reaching out to anyone online.
“Or try and talk to those people because one of the things I learned early on is you can talk to anyone on the Internet,”
-
Dylan Patel
Patel says you can reach out to anyone on the Internet and most say yes, but few people try.
“you can talk to anyone on the Internet, and you can just reach out to anyone. Yeah. And most people just say yes, but no one does it.”
-
Dylan Patel
Nadella describes connecting multiple Fairwater facilities on a petabit network spanning to Milwaukee for distributed AI training.
“Fairwater 4, which you're going to see under construction nearby, will also be on that one petabits network so that we can actually link the two at a very high rate.”
-
Dylan Patel
Patel says OpenAI shifted from InfiniBand to Ethernet for Stargate projects, following Meta's similar transition.
“All of OpenAI's cluster deployments in '23 and '24 were pretty much all InfiniBand. Now with their Stargate projects, they're all ethernet. And the same applies to other folks.”
-
-
-
Bill Gurley
Gurley argues Silicon Valley offers daily networking events that accelerate learning and connections by 10X or more.
“if you're in Silicon Valley and you have time, you can find an event every night. And at that event, you might meet a peer.”
-
Bill Gurley
Gurley argues Silicon Valley offers daily networking events that accelerate learning and connections.
“if you're a young energetic person at Silicon Valley, you can schedule something every night where you might meet something or learn something.”
-
Bill Gurley
Gurley says Silicon Valley has nightly networking events that massively increase probability of lucky career breaks.
“And on top of all that, what I just said, a lot of people talk about luck in their career path.”
-
Bill Gurley
Gurley says Silicon Valley's willingness to grant fifteen-minute meetings is shockingly higher than Wall Street.
“In Silicon Valley, the frequency with which you can ask someone for fifteen minutes of their time and get a yes is shockingly high.”
-
Bill Gurley
Gurley contrasts Silicon Valley's willingness to grant fifteen-minute meetings with other regions.
“I've never, I was always so surprised at how quickly people would take a fifteen minute meeting at Silicon Valley, like just compared to maybe some other places.”
-
Dylan Patel
Patel says networking represents 15-20% of cluster cost but has outsized impact on performance.
“with GPU rental pricing soaring, networking is 1520% of the cost of the cluster, but it can have a much more outsized impact on performance.”
-
Dylan Patel
Patel identifies disaggregated PD and wide EP as networking techniques where performance matters significantly.
“When someone has really good networking with these leading edge techniques to run a model such as KIMI or DeepSeq called disaggregated PD or wide EP, these are two different techniques, network performance is a humongous factor.”
-
Dylan Patel
Patel claims network performance drives 5x to 10x difference in inference for models using disaggregated PD or wide EP techniques.
“And so if you look into inference x, it turns out network performance drives a not just a, hey, 30%. It's actually multiple, like, x's.”
-
Bill Gurley
Gurley says human relationships and strong peer and mentor networks will matter way more in AI era.
“A lot of people have a theory that I think is reasonable that human relationships will matter way more. So, do you have a strong peer network? Do you have a strong mentor network?”
-
Bill Gurley
Gurley advises seeking mentors two steps above you rather than shooting too high like CEOs.
“the person that's two steps above you on the ladder has never been asked to be a mentor, and that's probably your best place to start, and they'll be flattered that you ask.”
-
Dylan Patel
Patel contrasts Nvidia's NVLink connecting 72 GPUs via switches versus Google's ICI connecting 8,000 chips without switches.
“For Google, their ICI can connect 8,000 chips at super high bandwidth, but you have to pass through other chips to get there because there's no switch.”
-
Bill Gurley
Gurley says when people are fascinated by something, others naturally connect them with opportunities.
“So if you knew someone who was just wouldn't stop chatting about modern architecture or whatever, and then you run into someone else and you go, oh, you should talk to Sally like this.”
-
Bill Gurley
Gurley bought Palm Pilot at conference to spam influential tech leaders at 70 cents per contact.
“People weren't really doing, like, cost of customer acquisition back then, but I think it was, like, 70¢ a name or something like that of the most influential people in the entire tech industry.”
-