Is open source AI really ahead of the frontier? 3 builders weigh in. | E25

August 6, 2026

AI Summary

5 min read

The open source frontier is closing in

When Jason Calacanis says he's been using only open source models for the past month and "can't tell the difference for 99% of my work," it's not just a personal anecdote. It's a signal that something fundamental has shifted in the AI landscape. The gap between frontier models from companies like OpenAI and Anthropic and the best open weight models has narrowed to the point where for most business tasks, the distinction is effectively invisible. And the implications are rippling through enterprise purchasing decisions, government policy, and the strategic calculus of every company building with AI.

The efficiency frontier is moving faster than the capability frontier

The panelists converged on a framework that explains what's happening. Eiso Kant from Poolside AI described it as a triangle of quality, cost, and speed. What's changed over the past year is that capabilities that once required an order of magnitude larger model can now be delivered by models an order of magnitude smaller. This isn't a one-time event—it appears to be happening on something like a yearly cycle.

Continue reading the full summary in the app — free to try.

Read Full Summary →

Free • No credit card required

Listen to Audio Summary Open in App

Never miss an episode of This Week in AI

Get every new episode summarized in your inbox — free, ~5 minutes to read.

No spam. Unsubscribe anytime.

What you'll learn

1 Timestamped Outline
2 (00:00) **Opening & Panel Introductions** - Host welcomes three builders: Iso Kant (Poolside AI, frontier coding models), Alex Cheema (Exo Labs, local AI clustering), and Alex Elias (Clue, AI recommendation engine)
3 (02:12) **Alibaba's Qwen3.8 Max and the Open-Source Price War** - Discussion of the 2.4 trillion parameter model that rivals frontier models at a fraction of the cost
4 (07:25) **The Triangle of Quality, Cost, and Speed** - Panel debates whether small specialized models will replace large frontier models or coexist with them
5 (09:55) **Why Coding is the Path to AGI** - Iso unpacks Poolside's thesis that coding is a proxy task for general intelligence, not the destination itself
6 (12:43) **Local.ai: New Benchmarks for Open-Weight Models** - Alex Cheema previews a benchmarking platform launching in a few days, designed specifically for models running locally
7 (16:25) **White House Shifts on Open Source** - The New York Times reports the administration will promote American AI models rather than block overseas open-weight models, sparking debate about national security vs. competitiveness

+ Full timestamped outline available in the app

Guests on this episode

Jason Calacanis Eiso Kant

Show Notes

This Week In Startups is made possible by:

PAYPAL OPEN

Today’s show:

China’s Alibaba just dropped a 2.4 trillion parameter open weight model that the company claims only trails Anthropic’s flagship among the major frontier rivals. Of course, this happened the very same week that the US government appears to have backed off from banning these models stateside entirely.

Are these models anywhere close to actually catching up to the latest and greatest releases from US labs like Anthropic and OpenAI? Really seems to depend on who you ask. (Even long time friends JCal and Elon Musk can’t see eye to eye on it.)

So on TWiAI Ep 25, we welcomed on Poolside AI CTO Eiso Kant, EXO Labs co-founder Alex Cheema, and Qloo founder Alex Elias to break down just how close Chinese labs are to catching up with the biggest and best models… and how long it will take US open weight builders to catch up.

PLUS we explore the cutting edge of AI-powered baby sleep monitors, consider Hank Green’s public AI apology, and dig into sci-fi author Ursula K. LeGuin’s “hand-mind” thesis.

Guests:

Eiso Kant on X: https://x.com/eisokant

Poolside: https://poolside.ai/

Alex Elias on X: https://x.com/ape

Qloo: https://www.qloo.com/

Alex Cheema on X: https://x.com/alexocheema

EXO Labs: https://exolabs.net/

Local AI benchmark platform: https://local.ai/

Relevant Links

Qwen3.8-Max launch announcement: https://qwen.ai/blog?id=qwen3.8

Bloomberg on Qwen3.8-Max launch: https://www.bloomberg.com/news/articles/2026-08-03/alibaba-drops-another-china-ai-model-with-breakthrough-performance

NYT: “White House Whipsaws Silicon Valley Over AI Rules”: https://www.nytimes.com/2026/08/04/technology/ai-washington-regulation-whiplash.html

Nvidia: “Open Weights and American AI Leadership”: https://images.nvidia.com/pdf/Open-Weights-and-American-AI-Leadership.pdf

World Bank: World Development Report 2026: https://thedocs.worldbank.org/en/doc/1e4e52502104a331fb42cba0d4afa995-0050062026/original/WDR2026-Concept-No