Connect your Claude, Codex or OpenClaw for full transcripts on millions of podcasts. Two week free trial, cancel anytime.

Start free trial →
AH Ash Hart On Intelligent Machines (Audio)

“If you're a big company, you're not bothered about the single stream. You're bothered about concurrency. Many users using that box to get as maximum throughput.”

Intelligent Machines (Audio) · Tech Strategy & Big Tech · November 2026

“If you're a big company, you're not bothered about the single stream. You're bothered about concurrency. Many users using that box to get as maximum throughput.” — Ash Hart, Intelligent Machines (Audio)

Context

Leo asked why the big AI labs never found the 2x to 6x single-user speedups his hobbyist engine gets. Ash's answer is that data-center operators optimise for total throughput across many customers, not one person's speed on their own hardware.

Listen to the episode → More from Ash Hart → More Tech Strategy & Big Tech quotes →

Transcript

Intelligent Machines (Audio) Around 12:57 into the episode
Paris Martineau

I actually... Why didn't

Leo Laporte

these, why didn't these companies do it themselves?

Ash Hart

It's quite easy to answer that. So if you're a big company, you're not bothered about the single stream. You're bothered about concurrency. Many users using that box to get as maximum throughput.

Leo Laporte

Because you're on a day in a data center where you're selling this to hundreds of thousands of customers.

Ash Hart

Exactly. So I applied this method last night. I actually got access to carbon. And I'm pretty impressed with it. That's amazing.

Leo Laporte

This is Google's new Gemini for. Yeah.

Speaker names come from our own diarization. The position is estimated from where the line sits in the episode.

More from Intelligent Machines (Audio)

“I just want you to remember that I said this on Veterans Day 2026. In one year, you're going to be amazed how cheap RAM is. Because none of these data centers are going to get built.” — Leo Laporte, Intelligent Machines (Audio) “You will find every other engine out there will employ weird mathematical equations to skip a few bits here, skip a few bits there, and it makes it quicker. But the output is not what the model would have given perfect serial decode, and that is what I couldn't lose. There's no point in having speed when there's quality degradation.” — Ash Hart, Intelligent Machines (Audio)