Open Source vs. Closed Source
TBPN

Open Source vs. Closed Source

July 27, 2026

Happy Monday.

We’re live on YouTube and 𝕏.

The current thing in tech and business is the open vs. closed source debate, spurred by Jensen Huang’s letter on open models.

Today’s LineupRun of Show

Open Source vs. Closed Source, by John Coogan

Alamy

Huge debate on the timeline this weekend over open source vs closed source AI.

Quick chain of events that got us here: April 7, 2026 Mythos / Project Glasswing showed really crazy cybersecurity capabilities. Then Fable came online from June 9–12 before going offline until July 1 onward. And just a few weeks later, Kimi K3 showed very near frontier abilities (especially good on front end benchmarks, sort of TBD on cybersecurity).

Distillation accusations flew immediately, and reports of lobbying to respond to Chinese open source quickly followed. The closed source frontier labs were pretty tight lipped on comms, but everyone sort of implicitly understands that having an open-source competitor model could potentially hurt growth and margins. There are also safety concerns, which people are deeply divided about.

A few years ago, the debate around open source centered on whether or not an open source ecosystem could actually flourish long-term. Would the capex requirements simply get too big to underwrite? Well we are certainly in that timeline now. Open source continues to deliver impressive models, so the question is around impacts both in business, geopolitics, and safety.

The market cap that has aligned behind the Nvidia open letter is huge (roughly $18T+) and that probably decides the issue. There won’t be a ban on open source AI, at least in the short term. But there are still other issues to work out. If open source labs are violating terms of service, what is the correct remediation path? Will the court system work effectively across borders in this case?

And then there’s the safety question. If open source AI increases the quantity and severity of cybersecurity incidents, how will that be offset? Will private companies just invest more in defense cybersecurity protections? Will governments do more to provide incident response during cyberattacks? Much to debate, so tune into the show today to watch us break it all down.

Clip Spotlight: Tyler Cowen says banning Chinese AI won't work

"Open source is coming and is here, whether we like it or not. Of course, a lot of it will be from China. We should allow American competitors to proceed on an even footing. But the attempt to outlaw it, ban it, or use sanctions against it is going to fail miserably.”

"I plead that the Trump administration gives up on this crusade. It will not work."

Headlines

Jensen Huang on why open models matter

Microsoft: Open Weights and American AI Leadership

WSJ: Nvidia Leads Defense of Open-Source AI With New Cybersecurity Initiative

Bloomberg’s Joe Weisenthal: The Four Words Everyone in Tech Should Say

WSJ: How to Beat China and Make AI Safe

Flo Crivello on 𝕏: We Must Ban Chinese Models

WSJ: Nvidia to Invest $5 Billion in Ilya Sutskever’s AI Research Lab

WSJ: Nvidia Bets on Ilya Sutskever’s New AI Lab to Expand Compute Reach

WSJ: Trump Caught Between Apple and Micron in Fight Over Chinese Chips

NPR: Paramount says it will pause its Warner acquisition while legal challenges play out

WSJ: The Mind-Bending Company That Gets a Million Job Applications—and Rejects 99.9%

FT: The new premium product: books written by people

WSJ: Big Companies Are Starting to Hire Again, Defying Predictions of AI Wipeout

Posts of the Day

Special thanks to our sponsors: Ramp, Shopify, CrowdStrike, MongoDB, NYSE, Codex, Public, Console, Railway, Figma, and Cisco.

Subscribe now

View on original site ↗