OpenvsClosedGet updates

Home / Inspired by

Inspired by the All-In podcast

This site began with an argument made on a podcast. Four hosts, four ways of seeing the same fight. Pick the door that sounds like you.

Open vs Closed is independent. It is not connected to, paid by or endorsed by the All-In podcast or its hosts. The summaries below are ours, in our words. Listen to the episodes to hear theirs.

Four doors

Four ways into the same story

All-In is a weekly show hosted by four investors. They rarely agree, and each one comes at open and closed AI from a different side.

The freedom argument

David Friedberg

Punish what people do with AI, not the AI itself. Rules that limit or label open software end up handing control to a few companies and the officials who oversee them.

On the show he called efforts to curb open AI “the path of tyranny”, and pointed to websites, browsers and smartphones as three times open software kept a technology in everyone's hands.

See the four futures

The ownership argument

Chamath Palihapitiya

A business that rents its AI hands its know-how to a supplier that could one day compete with it. The smart ones will want to own theirs.

On the show he has shown spending data suggesting more companies are buying their own machines to run open AI, and asked why America, with all its money, has not built a great open model of its own.

Open or closed for my business?

The monopoly argument

David Sacks

If one closed company wins everything, that is a monopoly, whatever it is called. The question is who is left to stand up for the open alternative.

On the show he has argued that the leading closed lab is building the most valuable monopoly in the world, and that it now describes its own market as winner-take-all.

See every law, and what it would control

The follow-the-money argument

Jason Calacanis

Ask who gains from each move. And remember what open means: nobody sells you an open AI. You copy it, run it on your own machines and can look inside.

As the show's host he connects each week's news: where the money and the talent went, and why closed labs could outbid everyone building in the open.

Follow the news

The story behind Door 1

It already happened once, with your web browser

On the show, David Friedberg points to three earlier fights that open software won: the software behind websites, web browsers and smartphones. The browser is the one everybody lived through.

2001

One company owned the door

Microsoft's Internet Explorer had roughly nine in ten users. With nobody to beat, it went five years without a major new version.

2004

A free copy anyone could improve

Firefox arrived, built on code Netscape had handed to the public. Nobody could buy it or shut it down. It passed 100 million downloads within a year.

Today

Nobody owns the door

Chrome, Safari and Firefox are all built on open software. You pick your browser, it costs nothing, and no company decides which websites you may visit.

When one company owned the way onto the internet, it stopped getting better. A free version that anyone could copy broke the lock. AI is at the same fork right now.

The browser story in full.

Episode library

Where to hear it for yourself

The episodes where open and closed AI came up, newest first, and what to listen for in each. We add to this list as new episodes air.

  1. E298 · November 2026

    Anthropic Goes Public, Nvidia in Crisis, Markets Tank, Open Source Crackdown?

    On the show: Calacanis, Palihapitiya, Friedberg

    • From 36:41: a reported White House plan to make companies disclose when they use Chinese open AI.
    • Friedberg's case that the answer is to punish misuse, not to limit or label software. This is the argument this site is built on.
    • Palihapitiya on why a company would rather own its AI than rent it.
    Find it on Apple Podcasts
  2. November 2026

    Jensen Ousted as Nvidia CEO: “Spy” Scandal, Export Restrictions, Market Impact

    On the show: Calacanis, Palihapitiya, Sacks, Friedberg

    • Friedberg on the two paths: AI that is widely available, or AI controlled by a few.
    • From 1:06:25: Sacks on a closed lab describing a winner-take-all market, and who will defend open AI now.
    Find it on Apple Podcasts
  3. E295 · October 30, 2026

    Anthropic S-1 Shocker, Greatest Business Ever? Google's $500B Wipeout, Midterm Predictions

    On the show: Calacanis, Palihapitiya, Sacks, Friedberg

    • From 32:47: why good-enough open AI at a low price is named as the biggest threat to a closed lab's business.
    Listen on Apple Podcasts
  4. October 2026

    AI Cyber Panic, Ken Griffin's Bubble Warning, The Fall of France?

    On the show: Calacanis, Palihapitiya, Friedberg, with guest Nikesh Arora

    • Palihapitiya's chart of companies buying their own machines to run open AI.
    • Arora, who runs a large security company, on preferring open AI because his team can inspect how it reaches an answer.
    Find it on Apple Podcasts

Episode titles and chapter times are the show's own. The notes on what to listen for are ours.

Find out when the future shifts

One short email when a law changes or one of the four futures gets closer. Nothing else.

Want to weigh in? Build a letter to your representative.