Issue No. 45Week ending Sunday, November 8, 2026803 episodes · 3,510 write-ups
The Podcast Summary.

The best ideas from 30+ top podcasts, every week.

AI infrastructure and compute

Dwarkesh Patel on AI infrastructure and compute

Host, Dwarkesh Podcast

4 quotes from 3 episodes on Dwarkesh Podcast, each with a timestamped link to the source.

4 quotes3 episodes

The short version

Dwarkesh Patel argues that physical hardware availability constrains AI compute scaling, even as agent swarms find ways to train models faster. ASML builds fewer than 100 extreme ultraviolet lithography tools each year, restricting global wafer output growth to just 2x annually.

Most interesting insights

A 50% API price increase at 200,000 tokens exposes memory bandwidth restrictions inside the physical cluster hardware.

“With longer context, Gemini 3.1 is 50% more expensive if you go over 200k tokens than if you're below 200k tokens.”

Dwarkesh Patel, Dwarkesh Podcast · April 2026 · Watch at 1:33:04 ↗

From LLM API Costs: Context Length Reveals Memory Bottlenecks

Long context windows will allow models to fully automate the work of top AI researchers by 2028.

“Here's why I don't think that the current lack of continual learning in these models will stop full automation of the best AI researchers by 2028…”

Dwarkesh Patel, Dwarkesh Podcast · November 2026 · Watch at 34:48 ↗

From Dwarkesh Patel on Automated AI Researchers by 2028

Top talking points

  1. Physical machinery shortages restrict global compute capacity

    AI compute demand grows 3x each year, while equipment availability restricts progress. ASML manufactures under 100 extreme ultraviolet lithography tools annually, capping leading-edge wafer output growth at roughly 2x.

    “For this EUV tool, ASML produces fewer than 100 of them a year.”

    Dwarkesh Patel, Dwarkesh Podcast · October 2026 · Watch at 48:20 ↗

    From ASML Tools Cap AI Chip Scaling at 2x

  2. Labs will distribute daily training runs to agent swarms

    Swarms of agents test smaller changes to bypass physical hardware constraints. By late 2027, top companies could provide enough daily compute for 10,000 individual agents to each run GPT-3 scale experiments.

    “And if this trend continues, then by the end of 2027, OpenAI could almost give 10,000 agents, so the amount that were involved in solving the Millennium Prize problem, each enough compute to run a GPT3 size training run every single day.”

    Dwarkesh Patel, Dwarkesh Podcast · November 2026 · Watch at 40:39 ↗

    From Dwarkesh Patel on Automated AI Researchers by 2028

Key takeaways from these write-ups

Dwarkesh Patel on Automated AI Researchers by 2028

  • Dwarkesh Patel forecasts that automated AI researchers capable of matching the top human researchers will arrive by 2028, unlocking recursive self-improvement.
  • Long context windows spanning multi-month working memory remove the need for constant architectural weight updates to achieve continual learning.

ASML Tools Cap AI Chip Scaling at 2x

  • ASML produces fewer than 100 extreme ultraviolet (EUV) lithography tools per year and will not expand far beyond that mark by 2030.
  • While AI compute demand grows 3x each year, global leading-edge wafer output can grow only about 2x in total, even with unlimited capital expenditure.

LLM API Costs: Context Length Reveals Memory Bottlenecks

  • Gemini 3.1's 50% price jump for context lengths over 200,000 tokens isn't arbitrary; it signals a hard constraint on memory bandwidth, not just compute, in underlying hardware.
  • Input (prefill) tokens costing up to 5 times less than output (decode) tokens indicates that the decode phase of LLM inference is heavily bottlenecked by memory bandwidth, not floating-point operations.

How we attribute quotes. Every quote was matched against the episode transcript, so the words and the timestamp are real (we trim filler words like "um", nothing else). The name comes from our written summary of the episode, and we use it only when a separate check of the captions finds that person on the episode. YouTube gives us no voice-by-voice transcript, so open the timestamp to hear who is talking. See a wrong name? Tell us and we fix or remove it.

More

The Sunday Email

Get next Sunday's issue in your inbox.

The week's best AI, tech, business, science, and culture podcasts in one short read. Free, every Sunday.

You get both newsletters, The Throughline and The Carry. No spam. Unsubscribe with one click.