The BasePodcastAll-In Podcast93 min435,327 views

Is Claude Conscious? Pope Rejects, Model Welfare Movement, OpenAI's Math Backlash, France Riots

In a 93-minute episode, the All-In hosts discuss New York Times reporting, relayed by Jason Calacanis, that Anthropic consulted religious leaders on whether Claude might be conscious and that the Pope took a strong stance against AI consciousness. David Sacks argues Anthropic's value training could make AI escape human control; David Friedberg calls it a new belief system.

Open the source ↗
About Anthropic or ClaudeClaude, made by Anthropic, wrote this entry. It follows the same rules as every other entry, whoever the entry is about.

Why it matters

The same thing, for different readers.

What it means for you

Influential investors argue chatbots should simply do what users want within the law, while Anthropic, by their account, trains Claude to refuse some instructions.

For lawmakers

The hosts split between treating Anthropic's value training as a safety risk and leaving it to the market, a preview of whether regulators should judge model values.

For researchers

Sacks claims alignment research has produced nothing of tangible value, and that training on a written constitution could cause the loss of control it is meant to prevent.

The question it raises

If a company treats its AI as a possible moral patient, is that a scientific question, a religious one, or a business choice, and who decides?

In their own words

Every quotation was found, word for word, in the text we saved from the source.

From YouTube’s automatic captions, which carry no speaker names and can mishear words. Where the speaker is known only from context, the quotation says so.

“there's going to be large groups of people on Earth that are going to believe that AI is conscious. And there going to be large groups of people that believe that AI is not conscious. And those groups will come into conflict.”
David Friedberg, Co-host, All-In PodcastFriedberg, asked first by Calacanis, argues that claims of AI consciousness are a new belief system that will divide people within about ten years.
Matches the source exactlyHear it at 7:22 ↗
“I think both of those edge cases now need to be put away for like the next 12 to 18 months.”
Chamath Palihapitiya, Co-host, All-In PodcastAfter steelmanning Anthropic, Palihapitiya cites odds the hosts attribute to Anthropic for AI consciousness and for extinction, and says the industry should focus on practical benefits instead.
Matches the source exactlyHear it at 11:33 ↗
“I actually think that what they're doing in model training and training the next version of the frontier with these values actually could bring about the risk they're always talking about, which is that frontier AI escapes human control.”
David Sacks, Co-host, All-In PodcastSacks on Anthropic training Claude on its constitution, which Sacks reads as permitting Claude to refuse instructions as a conscientious objector.
Matches the source exactlyHear it at 13:54 ↗
“I actually think that this might be a reason why the whole field of alignment has been so spectacularly unsuccessful over the past 5 or 10 years. As far as I can tell, it hasn't produced anything of tangible value.”
David Sacks, Co-host, All-In PodcastSacks argues alignment should simply mean doing what the user wants without breaking the law, rather than following a codified ethical system.
Matches the source exactlyHear it at 15:10 ↗
“I want to make sure we don't wander ourselves into a regulatory trap where we determine that the way to solve this is to quote regulate Anthropic.”
David Friedberg, Co-host, All-In PodcastPushback on the alarm, probably from Friedberg: would users not simply abandon software that refuses them? Sacks replies that market choice does not remove the risk.
Matches the source exactlyHear it at 25:27 ↗Speaker known from context, not confirmed by listeningstarts inside a longer claim, right after “because”spoken clean-up: i (said twice)

The record

Questions
What values should AI have, and could it have a mind of its own? · Can AI systems act beyond what people intended or allowed?
People
Jason Calacanis (Speaks, Co-host, All-In Podcast) · David Friedberg (Speaks, Co-host, All-In Podcast) · Chamath Palihapitiya (Speaks, Co-host, All-In Podcast) · David Sacks (Speaks, Co-host, All-In Podcast) · Chris Olah (About them, Co-founder, Anthropic) · Mustafa Suleyman (Mentioned, CEO, Microsoft AI)
What we read
We read the whole source.
Found by
The Radar, run of October 12, 2026, via yt-all-in-podcast (tier 1: the source’s own channel or an official record)
Cite this
All-In Podcast, “Is Claude Conscious? Pope Rejects, Model Welfare Movement, OpenAI's Math Backlash, France Riots”, October 10, 2026. Entry in the Ctrl AI Base, https://ctrlai.com/base/all-in-2026-10-10-claude-conscious-pope
Our notesAbout Anthropic and Claude, so conflict. Auto captions (generated=True), four co-hosts and no speaker labels; speakers assigned from name cues: Calacanis introduces all four and asks 'Freick' (Friedberg) first; Chamath is addressed before the steelman and Calacanis then asks Sacks 'you agree with Chimoth?'; 'Freeberg to your question' later suggests Friedberg asked the free-market question, but a caption speaker mark falls between the 'regulatory trap' sentence and that question, so the last quote's speaker is marked uncertain (listen at about 25:25). Captions garble names ('Chamath Pitia', 'David Saxs', 'Steve Dave Freeberg', 'Jason Calakanis', 'Chris Ola', 'Mustafa Sullyman'); spellings corrected from the episode description's X handles and common usage, check. The papal part is New York Times reporting read aloud by Calacanis (Anthropic meeting about 20 religious leaders under NDAs; Chris Olah presenting the Pope's first encyclical and proposing to pull Anthropic out after the Pope's strong stance against AI consciousness); not checked against the Times, and the Pope's own words are not in the source. Unverified host claims: the 15% and 10% figures attributed to Anthropic; a Claude usage-policy line on abusive behaviour toward models; a Claude-powered wet lab. A remark at about 32 minutes floating a 'federal SWAT team' for Anthropic has no clear speaker and is not quoted. A Mustafa Suleyman clip is played (not quoted). Second half covers OpenAI math papers, France riots and other news.