Jonathan Ross

It was a pleasure being back on 20VC with Harry Stebbings. His craft of interviewing is second to none and we went deep. This is the interview after we just launched 19,000 LPUs in Saudi Arabia. We built the largest inference cluster in the region. Link to the interview in the comments below!

When you make compute cheaper do people buy more? Yes. It's called Jevons Paradox and it's a big part of our business thesis. In the 1860s, an Englishman wrote a treatise on coal where he noted that every time steam engines got more efficient people bought more coal. 🧵(1/5)

Over the last 60 years, almost like clockwork, every decade compute gets about 1000x cheaper, people buy 100,000x as much of it, spending 100x times more overall.  Our mission at Groq is to drive the cost of compute towards zero;The cheaper we make compute the more people spend. 🧵(4/5)

When you make compute cheaper do people buy more? Yes. It's called Jevons Paradox and it's a big part of our business thesis. In the 1860s, an Englishman wrote a treatise on coal where he noted that every time steam engines got more efficient people bought more coal. 🧵(1/5)

It’s a paradox because if they're more efficient, why are they buying more coal? 🧵(2/5)

Groq just got a shout out on the All-In pod as one of the big winners for 2025 alongside Nvidia. It’s the year of the AI chip and ours is the fastest 😃

(1/5) One of the reasons why chips are so hard to innovate in is because if you're asking someone to put up a 10 million, 100 million, or a billion dollar check they need to know that what they're buying is going to work.

(4/5) We're not shipping anyone millions of dollars of hardware. It’s not a big ask for them to try it. And when they try it, they love it.

(1/5) One of the reasons why chips are so hard to innovate in is because if you're asking someone to put up a 10 million, 100 million, or a billion dollar check they need to know that what they're buying is going to work.

(2/5) That makes it almost impossible to do counterintuitive things. Like, “Try this new chip called an LPU", when everything in the zeitgeist is talking about GPUs. And if you're a startup? Forget it. That realization is why we made the strategic decision to put up our own cloud.

(1/5) "The reports of the LLM scaling laws' demise have been greatly exaggerated." techcrunch.com/2024/12/06/m...

(4/5) It's also significantly less expensive and faster than the larger model. Meta continues to push the lead in open weight innovations, and is keeping the pressure high for proprietary model providers to attempt to keep ahead of the giant wave of open.

(1/5) "The reports of the LLM scaling laws' demise have been greatly exaggerated." techcrunch.com/2024/12/06/m...

(2/5) Today, our partner Meta released its latest version of Llama-3.3-70B-Instruct. And to all those who speculated that the industry had hit the wall - maybe some have, but Meta hasn’t yet. 😉 This is a big deal.

(1/5) The question I get asked a lot is, “Should I be afraid of AI?” There was this guy who got in a lot of trouble once, his name was Galileo.

(4/5) When you look at something that's rapidly becoming that grand, that vast, and that capable it makes you feel small and afraid. Over time we've gotten used to this idea that the universe is large and that we're small. If anything we now see the beauty in how vast the universe is.

(1/5) The question I get asked a lot is, “Should I be afraid of AI?” There was this guy who got in a lot of trouble once, his name was Galileo.

(2/5) He got in trouble because he made a telescope, pointed it at the stars, and using data from other astronomers he put together a pretty convincing case that human beings weren't the center of the universe. They locked him in a tower.

(1/5) Everyone at Groq has one of these challenge coins on them. It’s how we create alignment. One side says its 25 million, because we're going to get to 25 million tokens per second by the end of the year On the other side, it says, “Make it real. Make it now. Make it wow.”

(4/5) The reason for that number? 25 million? If we do 25 million tokens per second then we're going to have as much compute capacity as a hyperscaler had at the start of this year. And from that point forward, we'll just ramp.

(1/5) Everyone at Groq has one of these challenge coins on them. It’s how we create alignment. One side says its 25 million, because we're going to get to 25 million tokens per second by the end of the year On the other side, it says, “Make it real. Make it now. Make it wow.”

(2/5) Okay fine, cute prop. But how does a challenge coin create alignment? Well, if someone makes a proposal and it doesn't help with these priorities, anyone can just put their coin on the table and ask “how does it help with this?”