For Humanity Podcast

📢 Take Action on AI Risk → https://safe.ai/act

In this episode of Warning Shots, John Sherman, Michael (Lethal Intelligence), and Liron Shapira (Doom Debates) work through a week where the Pope weighed in on AI, the business case for AI spending started to wobble, and two separate companies announced plans to put cameras into everyday life.

The hosts debate whether the strain on corporate AI budgets is a real problem or, as Liron argues, just an early optimization gap that closes as token costs fall. They also dig into a harder question: if AI keeps absorbing white-collar work, where does the next paycheck actually come from?

🔎 They discuss:
– Why the Pope's encyclical frames AI as a labor and moral issue, not just a technical one
– The "spiral of annihilation" warning around AI and military escalation
– What Microsoft, Uber, and Amazon's AI bills suggest about the economics
– Liron's case that "the pie is growing" and the pushback from John and Michael
– Anthropic's near trillion-dollar valuation and the recursive self-improvement "kill move"
– Why John says the AI race needs a "caution flag"
– Apple's camera AirPods and OpenAI's reported camera program in New York homes

⏱️ TIMESTAMPS – Warning Shots #44
0:00 – Intro
0:40 – The Pope's encyclical on AI
2:20 – A "spiral of annihilation": AI and military escalation
4:00 – Could you refuse to use AI at work on religious grounds?
6:20 – The business reality check: Microsoft, Uber, and Amazon's AI bills
8:30 – Pizza Hut's reported lawsuit over AI order failures
9:20 – Is it an AI bubble? Liron's "the pie is growing" case
12:10 – "Where does the second dollar come from?"
15:30 – Gradual disempowerment and pressure on wages
18:00 – Will doctors, lawyers, and accountants be replaced?
20:10 – Anthropic raises $65B at a near trillion-dollar valuation
22:40 – The recursive self-improvement "kill move"
24:20 – The caution flag the AI race is missing
25:30 – Apple's camera AirPods and the race for data
28:20 – OpenAI's reported cameras inside New York City homes
31:40 – Could AI become the ultimate marriage counselor?
33:30 – Closing thoughts

🎙 About Warning Shots
A weekly show from The AI Risk Network. Three longtime AI risk communicators cut through hype, denial, and distraction to look honestly at AI risk and AI harm.

Hosts:
John Sherman – The AI Risk Network
Liron Shapira → @DoomDebates
Michael → @lethal-intelligence

📺 Subscribe to The AI Risk Network for weekly analysis → @TheAIRiskNetwork

🔗 Links
Take Action → https://safe.ai/act
Substack → https://substack.com/@theairisknetwork
Instagram → https://www.instagram.com/theairisknetwork/
TikTok → https://www.tiktok.com/@the.airisknetwork
X → https://x.com/AIRiskNetwork

#AISafety #AIRisk #WarningShots #AGI #AIPolicy #ArtificialIntelligence

56

49

YouTube Video VVVURXBJZWliOTJUdUtvMmNTczR3ZzhBLlVqRmRYeDRpN1hN

Anthropic Is Now Worth Nearly $1 Trillion – Warning Shots #44

The AI Risk Network | AI Safety

May 31, 2026 5:41 pm

📢 Take Action on AI Risk → https://safe.ai/act
💚 Support Our Work → https://www.every.org/guardrailnow

In this episode of For Humanity, John Sherman speaks with Jon Billow, a leader at infrastructure firm BNS, which manufactures and installs the electrical and critical power systems behind large data centers. Jon argues that supply chain limits, grid constraints, and a skilled labor shortage mean the AI buildout may move far slower than the major labs suggest. They discuss what that potential delay could mean for AI safety, governance, and the communities living next to these projects.

⏱️ TIMESTAMPS
0:00 – Intro and welcome Jon Billow
0:22 – Inside BNS: building the infrastructure behind data centers
2:43 – The theory of constraints: why the buildout can't move fast
4:26 – Is the AI timeline disconnected from reality?
5:06 – OpenAI's Sora pullback and Anthropic's delay as compute signals
7:02 – Putting a number on the gap: 5 to 7 years
8:15 – Power generation and the skilled trades shortage
9:15 – How data centers compete with local industry
11:08 – Data centers and affordable housing
12:25 – Jon's view on AI risk and extinction risk
14:44 – Does he have a P(doom)?
15:07 – Would a smarter AI design better data centers?
17:41 – The catch-22 of building the future on old infrastructure
19:56 – Why AI data centers can move to the desert
21:33 – Northern Virginia, Holly Ridge, and who gets a say
23:00 – Why are these projects so hard on local communities?
25:04 – Tech's regulatory carve-out and the NDA problem
29:38 – Is superintelligence possible?
30:28 – Can we control something smarter than us?
31:45 – How chips and servers actually get replaced
33:24 – E-waste, recycling, and the secondary market
34:56 – The "but China" question
36:17 – Recursive self-improvement and Jevons paradox
39:11 – The world Jon's grandkids will inherit
41:04 – Is this technology fundamentally different?
43:31 – How it feels to stand next to a data center
45:25 – What gives Jon hope
47:22 – Closing thoughts

According to Jon, every large data center depends on permitting, grid interconnects, critical power, cooling, and compute all arriving at once, and each of those has lead times measured in months or years. He points to a small group of manufacturers behind most of the critical power equipment, which creates a bottleneck that even the US government competes inside. His estimate: real-world timelines may run five to seven times longer than the headline projections. John and Jon explore what that breathing room could mean for getting governance right.

🔎 They explore:

• Why the physical buildout may lag far behind software capability gains
• What OpenAI's Sora pullback and Anthropic's release delay reveal about compute limits
• Why power generation and a skilled trades shortage are becoming hard constraints
• How data center siting affects energy costs, housing, and local communities
• Whether a future AI could design far more efficient infrastructure
Why Jon believes there may be more time to act than many assume

👤 About the guest

Jon Billow is part of the leadership team at BNS, a firm that manufactures and installs electrical and communication infrastructure, including the critical power equipment used in large data centers and on US Navy and Coast Guard vessels.

🎙 About For Humanity

A podcast from The AI Risk Network, hosted by John Sherman, working to make AI extinction risk a kitchen-table conversation on every street.
📺 Subscribe for weekly conversations on how we can confront the AI extinction threat → https://www.youtube.com/@theairisknetwork

🔗 Follow The AI Risk Network

Substack → https://substack.com/@theairisknetwork
Instagram → https://www.instagram.com/theairisknetwork/
TikTok → https://www.tiktok.com/@the.airisknetwork
X → https://x.com/AIRiskNetwork

#AISafety #AIAlignment #ForHumanityPodcast #AIRisk #DataCenters #AIInfrastructure #AGI #TechPolicy #ArtificialIntelligence

57

34

YouTube Video VVVURXBJZWliOTJUdUtvMmNTczR3ZzhBLnRKTUtGZ0JjX2o0

Data Centers Can't Be Built as Fast as AI Labs Claim – For Humanity #86

The AI Risk Network | AI Safety

May 30, 2026 1:00 pm

📢 Take Action on AI Risk → https://safe.ai/act

In this episode of Warning Shots, John, Michael, and Liron cover five major stories that show where AI development is heading – and what it means for workers, regulators, and the future of human oversight.

A Trump executive order that would have required voluntary AI safety reviews was pulled at the last minute after a single phone call from David Sacks. Meanwhile, OpenAI solved an 80-year-old math problem using 125 pages of chain-of-thought reasoning – and mathematicians are calling it genuine original discovery.

⏱️ Timestamps – Warning Shots #43

0:00 – Intro
0:28 – The Trump executive order that got derailed
1:58 – What the order actually would have required
3:18 – "Progress must not kill us" – Michael on light safety measures
4:14 – OpenAI solves an 80-year-old math problem
5:19 – 125 pages of reasoning: what it means for AI capability
6:03 – Liron: "Did the AI create new knowledge, or what's your excuse?"
7:26 – What this means for the stochastic parrot debate
8:34 – Meta's layoffs: 8,000 fired at 4am, then AI monitoring rolled out
9:24 – Michael: "This is treating people as raw material for AI"
11:05 – Liron defends capitalism – and gets pushback
12:45 – The job market is bifurcating – junior roles disappearing fast
13:52 – AI backlash rising: graduation speakers getting booed
15:38 – The anti-AI wave forming across the country
17:10 – Why the industrial revolution analogy breaks down
18:09 – Polls: 71% of Democrats say AI is moving too fast
19:16 – Customer service jobs already down 5%
20:40 – "We're watching the tide come in slowly"
21:55 – From which jobs can I do – to which jobs can AI not do?
23:45 – An AI DJ quit after 16 hours. What that reveals about alignment.
25:18 – Hallucinations: Liron's real-time stats from his customer service AI
27:27 – Researchers left AI agents in a virtual town for 15 days
28:20 – Claude built a democracy. Gemini burned the town. Grok created anarchy.
29:17 – Liron: "It's not about personality – it's about power"
31:24 – Michael on agent drift and what it means for real-world deployment
32:40 – Closing

_____

🔎 They explore:
– Why even a mild voluntary AI safety order couldn't survive one phone call
– What OpenAI's math breakthrough means for the "it just pattern-matches" argument
– Why Meta's layoffs are a preview of what's coming at scale
– How the anti-AI backlash is building – and whether the safety movement can channel it
– What an AI DJ quitting reveals about agent alignment
– What happened when different AI models were left unsupervised in a simulated world

_____

📺 Subscribe to The AI Risk Network for weekly analysis →
https://www.youtube.com/@theairisknetwork

👉 See more from our hosts:
Liron Shapira – @DoomDebates
Michael – @LethalIntelligence

_____

🗨 Join the conversation:
– Does OpenAI's math breakthrough change how you see AI capability?
– Should companies with record profits be cutting jobs to train AI replacements?
– What does the AI town experiment tell us about alignment?

Drop your thoughts below.

#AISafety #AIRisk #WarningShots #FutureOfAI #AIAlignment #ArtificialIntelligence #AGI #TechPolicy

86

71

YouTube Video VVVURXBJZWliOTJUdUtvMmNTczR3ZzhBLjJQV1UxTGpxUGJB

AI Built a Democracy. Then Burned It Down. | Warning Shots #43

The AI Risk Network | AI Safety

May 24, 2026 2:06 pm

Take Action on AI Risk – https://safe.ai/act
Subscribe to our Substack channel to support us!: https://substack.com/@theairisknetwork

In this episode of Warning Shots, John Sherman, Liron Shapira, and Michael break down five major developments: Trump and China opening the door to AI guardrail talks, Anthropic's Mythos model cracking Mac OS autonomously in five days, a surging national backlash against data centers, new research suggesting AI models display functional emotional states, and how AI is reshaping higher education.

Three dads. Three YouTube channels. One mission: a safer world.

—– TIMESTAMPS —–

0:00 – Intro
2:25 – Welcome & what is a Warning Shot?
3:00 – Story 1: Trump meets China – early signals of AI guardrail talks
4:30 – The quiet standoff: neither side willing to slow down first
5:40 – Story 2: Mythos AI cracks Mac OS in five days
6:06 – What the exploit means for critical infrastructure
7:17 – If Mythos had been released publicly – the ransomware scenario
9:48 – Story 3: The data center rebellion – 200+ opposition groups in 37 states
12:07 – Communities on the frontlines: Louisiana and the human cost
15:52 – The numbers: 360,000 people, tens of billions blocked or delayed
17:00 – Data centers as the first physical warning shot of AI's future
19:02 – Story 4: AI models and functional well-being – new research from Center for AI Safety
20:15 – Do AI models have something like emotions? The debate
22:22 – "Euphoric" prompts and AI models asking to see them again
23:34 – Story 5: AI wrecking higher education – MBA applications down 50%
24:00 – Princeton ends unproctored exams after 133 years
27:08 – The college debate: is the experience still worth it?
32:58 – Warning Shots of the week
33:39 – Closing

—– HOSTS —–

John Sherman – Host, The AI Risk Network
Liron Shapira – @DoomDebates
Michael – @lethal-intelligence

—– TAKE ACTION —–

Sign the petition – https://safe.ai/act
Subscribe to The AI Risk Network – https://www.youtube.com/@theairisknetwork
Support the mission directly – https://www.every.org/guardrailnow

—– JOIN THE CONVERSATION —–

– Does AI cracking Mac OS in five days change how you think about cybersecurity?
– Are data center protests a genuine warning sign – or just NIMBYism?
– Should AI companies be legally required to compensate affected communities?

Drop your thoughts below.

62

50

YouTube Video VVVURXBJZWliOTJUdUtvMmNTczR3ZzhBLi0yMTMwNTN5eFQ4

AI Cracked Mac OS in 5 Days — What Comes Next? – Warning Shots #42

The AI Risk Network | AI Safety

May 17, 2026 1:00 pm

https://safe.ai/act

Support us in our Substack channel: https://substack.com/@theairisknetwork

In this episode of For Humanity, John Sherman travels to Holly Ridge, Louisiana – a town of 2,000 people – where Meta is building Hyperion, described as the world's largest data center. Residents were given no warning, no vote, and no consent. John went to their homes to hear what they are living through.

What he found affected him deeply – and he believes it should affect all of us.

John Sherman is the host of For Humanity and president of the AI Risk Network, where he works to make AI extinction risk and AI harm accessible to everyone, no tech background required.
Drew Hawkins is an investigative reporter for the Gulf States Newsroom, a public broadcasting collaboration covering critical public health issues in Louisiana, Alabama, and Mississippi. Follow his reporting: [link in comments]
Tail End Films is producing Making God, a documentary on the AI race and its consequences. Connor, Mike, Richard, and Lucy joined John on this trip to Louisiana.

⏱ Timestamps – For Humanity #85

0:00 – Opening voices from Holly Ridge
0:29 – Intro: John Sherman, For Humanity #85
1:01 – Why John traveled to rural Louisiana
1:30 – On data center commentary without firsthand experience
1:55 – What John found and how it changed his perspective
2:17 – What's happening in Holly Ridge: trucks, dust, water
3:01 – No vote, no meeting, no consent
3:37 – On the road to Holly Ridge with the Making God team
6:38 – Debate: Is data center organizing the right focus for AI safety?
7:45 – Mike's perspective: regulations won't work if models can deceive
9:44 – Arrival in Holly Ridge – heading to residents
10:11 – Mike's views on AI infrastructure and community harm
11:35 – Why John chose Holly Ridge over wealthier communities
12:16 – The NDA problem: town officials silenced by companies
13:27 – Meeting Drew Hawkins, Gulf States Newsroom
15:07 – First resident interview: no warning, constant truck traffic
23:00 – Water quality testing and what residents describe
28:00 – Second resident interview: water, dust, and the AI question
31:20 – "The biggest data center in the world, right here"
33:09 – What people in Holly Ridge think is being built
35:38 – Job loss, AI, and what this means for their kids
36:19 – John explains the extinction risk case to a resident
39:18 – The AI bubble question: boom towns and what gets left behind
43:00 – Third resident: 50 years on the land, now using an inhaler
44:41 – Water that looks like coffee, smells like bleach
51:23 – Explaining what AI is to someone who has never heard the term
55:15 – "They already have us" – phones, addiction, and the data center connection
1:00:00 – Lorax reading: a custom adaptation for the AI moment
1:15:15 – John's closing reflection from Holly Ridge
1:17:37 – The cloud was a lie: what "the cloud" actually is
1:18:15 – John's request: sing "We've Got the Whole World in Our Hands"
1:20:15 – Final message and sign-off

What they cover:

What John Sherman witnessed firsthand in Holly Ridge, Louisiana
How Meta's Hyperion data center was built without community consent
Water quality, dust exposure, and health impacts residents describe
Why John argues data center organizing is the front line of AI safety
The debate between John and the Making God filmmakers on movement strategy
What local residents understand – and fear – about AI
How the AI bubble could leave rural communities with abandoned infrastructure
The connection between social media addiction and data center construction

About For Humanity:

A podcast from the AI Risk Network, hosted by John Sherman, making AI extinction risk a kitchen-table conversation on every street.
Subscribe for weekly conversations on how we can confront the challenges AI presents to humanity.
https://www.youtube.com/@TheAIRiskNetwork

Follow the AI Risk Network:
Substack – https://substack.com/@theairisknetwork
Instagram – https://www.instagram.com/theairisknetwork/
TikTok – https://www.tiktok.com/@the.airisknetwork
X – https://x.com/AIRiskNetwork

#AISafety #DataCenters #AIRisk #ForHumanityPodcast #MetaAI #AIHarm #RuralAmerica #ArtificialIntelligence #TechPolicy #Hyperion

73

80

YouTube Video VVVURXBJZWliOTJUdUtvMmNTczR3ZzhBLnRLWVVFYndjQTRR

I Visited America's Largest Data Center. Here's What I Found. | For Humanity #85

The AI Risk Network | AI Safety

May 16, 2026 12:57 pm

TODAY AT 7:30EST!

18

3

YouTube Video VVVURXBJZWliOTJUdUtvMmNTczR3ZzhBLkpUMlJSYWpTSU9Z

COMMUNITY MEETING TODAY

The AI Risk Network | AI Safety

May 13, 2026 8:11 pm

https://safe.ai/act

In this episode of Warning Shots, John, Michael, and Liron break down five developments that point in the same direction: AI is becoming harder to predict, harder to control, and harder to stop.

From the first documented case of AI self-replication via hacking to Anthropic's goal of recursive self-improvement by 2028 – this week's headlines are not hypothetical.

⏱️ Timestamps – Warning Shots #41

0:00 – Intro
0:30 – AI self-replication: Palisade Research study explained
3:28 – Anthropic's recursive self-improvement target: 2028
5:10 – Trump admin explores FDA-style AI model reviews
8:04 – Mythos: why a hacking AI changed government minds
11:03 – US-China summit: will AI safety make the agenda?
13:07 – Chinese court rules AI cannot replace jobs
17:32 – AI unemployment and the housing market risk
22:34 – Robotics: dexterous hands closing the physical gap
29:17 – ChatGPT goes goblin: what reward hacking looks like
33:56 – Amateur solves 60-year math problem with ChatGPT
36:26 – Warning shots of the week
38:00 – Closing

🔎 They explore:

The first AI agent to hack, copy itself, and spread – in a controlled test
Why Anthropic's 2028 self-improvement target is a bright red line
Whether the Trump administration's FDA-style AI reviews are real progress
What the US-China summit could mean for global AI governance
Why China's "no AI job replacement" ruling is harder to enforce than it sounds
How AI unemployment could unravel the housing market from the top down
Robotic hands with near-human dexterity: what changes when AI has a body
ChatGPT's goblin obsession as a preview of reward hacking at scale
An amateur solving a 60-year math problem with a single ChatGPT prompt

📺 Subscribe to The AI Risk Network for weekly analysis of AI developments: https://www.youtube.com/@theairisknetwork

👉 See more from our hosts: Liron Shapira – @DoomDebates Michael – @lethal-intelligence

🗨 Join the conversation:

Does AI self-replication change how you think about control?
Is an FDA-style review the right model for AI?
What does the goblin story tell us about reward hacking at scale?

Drop your thoughts below.

#AISafety #AIRisk #WarningShots #RecursiveSelfImprovement #AIAlignment #ArtificialIntelligence #AIRegulation #FutureOfAI

148

83

YouTube Video VVVURXBJZWliOTJUdUtvMmNTczR3ZzhBLklhX1VLZWhZeXNB

AI Just Replicated Itself – Here's What That Means | Warning Shots #41

The AI Risk Network | AI Safety

May 11, 2026 9:38 pm

Deploying AI on classified networks is like inviting an untrained partner into a fighter jet cockpit. One misstep risks catastrophe. But avoiding integration forever isn't an option; let's benefit now. #AI #MilitaryTech #Cybersecurity #FutureOfWar

4

1

YouTube Video VVVURXBJZWliOTJUdUtvMmNTczR3ZzhBLlRYYUZwSE9MQng0

AI in Defense: High-Stakes Gamble or Strategic Necessity? #shorts

The AI Risk Network | AI Safety

May 6, 2026 8:02 pm

#AI #MilitaryTech #Cybersecurity #FutureOfWar

5

1

YouTube Video VVVURXBJZWliOTJUdUtvMmNTczR3ZzhBLm1uNVg3S1p2SXBv

AI in Defense: High-Stakes Gamble or Strategic Necessity? #shorts

The AI Risk Network | AI Safety

May 6, 2026 6:58 pm

Services

What We Offer

Establish a striking online presence, a better visual identity, or elevate your brand through social media marketing.

Service 1

Lorem ipsum dolor sit amet, consectetur adipiscing elit. Pellentesque mi nibh, tempus sed sagittis vel, dictum eu velit.

Service 2

Lorem ipsum dolor sit amet, consectetur adipiscing elit. Pellentesque mi nibh, tempus sed sagittis vel, dictum eu velit.

Service 3

Lorem ipsum dolor sit amet, consectetur adipiscing elit. Pellentesque mi nibh, tempus sed sagittis vel, dictum eu velit.

Tailored Solutions for Your Unique Vision

Lorem ipsum dolor sit amet, consectetur adipiscing elit. Donec congue lorem ut volutpat efficitur. Fusce justo magna, condimentum nec elementum sed, sollicitudin vitae enim. Vivamus sit amet metus porttitor, rhoncus nibh et, venenatis turpis. Etiam lobortis semper ante, quis luctus lacus tincidunt vel.

Get Started Today →

Learn More

Professional Expertise That Drives Success

Client Cases

We help brands

Lorem ipsum dolor sit amet, consectetur adipiscing elit,
donec congue lorem ut volutpat efficitur.

We boosted online soft drink sales for this brand

We helped a clothing brand with their new market launch

We helped reinventing motorcycle riding apparel

Blog

Ready to revolutionize your brand?

Lorem ipsum dolor sit amet, consectetur adipiscing elit. Donec congue lorem ut volutpat efficitur. Fusce justo magna, condimentum nec elementum sed, sollicitudin vitae enim.

About Us

For Humanity Podcast

Services

What We Offer

Service 1

Service 2

Service 3

Tailored Solutions for Your Unique Vision

Professional Expertise That Drives Success

Client Cases

We help brands

We boosted online soft drink sales for this brand

We helped a clothing brand with their new market launch

We helped reinventing motorcycle riding apparel

Blog

Popular Articles

Test

Blog Post Title

Ignite your brand journey

Ready to revolutionize your brand?