Ajay Patel, Head of World ID at Tools for Humanity, joins Mark and Jeremy to discuss digital identity in the age of artificial intelligence.

As AI-generated content spreads across the internet and bots become harder to distinguish from people, existing verification systems are starting to fail. CAPTCHAs are increasingly ineffective, while conventional identity checks often require users to hand over sensitive personal information.

World ID proposes a different model: verifying that someone is a unique human without revealing their identity. The system combines iris biometrics, zero-knowledge proofs and decentralised infrastructure to create what Tools for Humanity describes as a privacy-preserving proof of personhood.

In this episode, we discuss:

• How Sam Altman’s World ID system works

• Why World ID uses iris scans to verify unique humans

• What zero-knowledge proofs mean for online privacy

• Whether biometric data can ever be collected safely

• How AI agents and bots are changing digital identity

• What happens when platforms can no longer tell humans from machines

The conversation examines the technical design, privacy concerns and wider implications of World ID. It also asks a more fundamental question: what kind of internet remains when proving you’re human becomes difficult?

Please enjoy the episode and share it with someone interested in AI, digital identity and online privacy.

–

Timestamps

(00:00) Intro: Future of Digital Identity & AI Challenges

(01:15) Meet Ajay Patel: Head of World ID at Tools for Humanity

(01:44) WorldID: Why Digital Identity Matters in the Age of AI

(05:40) Global Digital Identity for 8 Billion Humans

(07:38) AI in 2025: How Advanced is AI Really?

(08:17) AI Bypassing CAPTCHA: The Security Crisis

(12:54) ZK Proofs Explained: Privacy-Preserving Tech Behind World ID

(15:20) How World ID’s Iris Scan Works (No Data Stored!)

(18:54) World ID Data Security: Breaking Down Multi-Party Computation

(22:56) Hot Button Q&A: Apple vs Android, AGI Threats, & More

(24:19) World ID Use Cases: Gaming, Voting, & Anti-Bot Systems

(25:13) Exclusive: World ID x Razer to End Gaming Cheats

(30:56) AI Hype vs Reality: Are Bots Just “Yes Men on Servers”?

(35:35) Decentralizing World ID: Open Protocol, No Corporate Control

(38:15) Question for Next Guest: Hollywood’s David Bianchi on AI & Art

–

Thank you for Thinking On Paper. Your future self will thank you.   

In the final chapter of Nexus, Yuval Noah Harari argues that artificial intelligence could divide the world into competing digital empires, each built on separate hardware, software, data and political systems.

Mark and Jeremy examine Harari’s idea of a “Silicon Curtain” and what it could mean for the future of AI, geopolitics and the internet. Rather than producing one shared technological future, AI may accelerate the fragmentation of the global information system.

In this episode, we discuss:

• What Harari means by the Silicon Curtain

• How the United States and China are building competing AI ecosystems

• The role of data colonialism in the development of artificial intelligence

• How avatars, deepfakes and digital identities affect our understanding of reality

• Why digital personas may become more influential than physical individuals

• How software, data and computing infrastructure are reshaping geopolitical power

• Whether societies can build more reliable information networks in an increasingly divided world

  
  

This episode isn’t primarily about artificial general intelligence. It’s about who controls AI systems, how those systems shape perception, and whether a shared global information order can survive.

Please enjoy the show.

–

Chapters

(00:00) Disruptors and curious minds

(01:54) My AI Epiphany

(02:54) The precondition for cooperation isn't similarity. It's the ability to exchange information

(04:28) AI Pessimism - How many nefarious bad actors do you need

(05:50) The Liverpool Manchester Train moment for AI

(06:48) Google was an AI company in 2002

(08:21) Cats and consciousness

(08:55) ImageNet Large Scale Visual Recognition Challenge

(11:50) The mind-body problem and online identities

(15:26) Veneer theory and realism

(17:22) Data colonialism and digital "cocoons"

(21:58) Epilogue: Create wiser networks

(24:13) How to be wise in the age of AI

--

Read books with us: www.thinkingonpaper.xyz

In Chapter 10 of Nexus, Yuval Noah Harari examines whether artificial intelligence will strengthen authoritarian regimes or expose weaknesses within them.

Dictatorships rely on fear, loyalty and centralised control. AI feels no fear, has no political allegiance and can process information at a scale no human bureaucracy can match. That could make it an effective tool for surveillance and censorship. It could also create new problems for regimes that depend on controlling what their institutions know and say.

In this episode of the Thinking on Paper Book Club, Mark and Jeremy discuss AI, dictatorship and the structure of authoritarian power.

We cover:

• How authoritarian governments could use AI for surveillance and censorship

• What happens when an AI system identifies contradictions in the ideology it is trained to enforce

• How algorithmic censorship could complicate attempts to erase people and events from history

• Why dictatorships struggle with complexity and accurate feedback

• How information failures weaken totalitarian systems

• What the Russell-Einstein Manifesto can teach us about the risks of AI

• Whether machines could destabilise the political systems they were built to protect

AI could give authoritarian governments more control over citizens, institutions and information. But it could also make their internal contradictions harder to contain.

This episode asks whether AI will become the most powerful instrument of dictatorship yet, or a source of instability that authoritarian leaders cannot fully control.

Please enjoy the book club. 

–

CHAPTERS

(00:00) Disruptors and curious minds

(00:21) AI spectrum of doom - utopia

(02:10) Damnatio memoriae

(04:28) The Alignment Problem Russian Style

(06:45) Self correcting mechanisms for totalitarian nut-jobs

(09:05) Super Machiavellian AI

(09:51) The Russell-Einstein Manifesto

–

Read books with us - www.thinkingonpaper.xyz

Like, subscribe, tell a friend.

Diego Borgo has helped develop Web3 strategies for brands including Adidas, Porsche and Mastercard. In this episode, he explains why many corporate blockchain and NFT projects failed, and what brands should do differently.

Much of Web3 marketing prioritised hype, technical jargon and short-term attention over clear customer value. Brands launched tokens, NFT collections and online communities without establishing why their audiences should care or how the technology supported a broader business strategy.

Diego joins Mark and Jeremy to examine what went wrong and how marketers can approach emerging technologies more effectively.

We discuss:

• Why many Web3 marketing campaigns failed

• How brands mishandled customer onboarding

• What makes technology-led communities valuable and sustainable

• Why artificial scarcity and superficial engagement rarely build trust

• Diego’s “See the Unseen” framework for identifying market shifts

• Where AI and blockchain can create practical business value

• What marketers misunderstand about decentralisation, culture and ownership

• How executives should evaluate Web3 projects before launching them

This episode is a practical discussion about Web3 marketing, brand strategy and emerging technology. It’s intended for marketers, strategists and executives deciding whether blockchain, AI or digital ownership belongs in their business.

Please enjoy the show.

--

Follow Diego: https://x.com/don_borgo

 --

Chapters

(00:00) Disruptors And Curious Minds

(00:35) The Problem With Web3

(01:56) Welcome To The Show Diego Borgo

(02:19) Simplifying Web3 for Your Family

(06:14) Audience Infrastructure in Web3 - What Did Brands Get Wrong?

(09:52) Short-Term vs Long-Term Brand Strategies

(12:31) The Web3 Echo Chamber

(14:53) Meet People Where They Are

(16:56) Web3 On-boarding Is Patronizing

(17:58) Hot Buttons

(19:04) What Has Diego Changed His Mind About Regarding AI

(23:08) Decentralized AI

(28:19) Trust 

(30:31) Follow On Question: AI And Blockchain Scaling

(32:54) UBI And Diego's Question For WorldID

(35:11) Backstage After Show

--

Full episodes and archives: www.thinkingonpaper.xyz

Watch On YouTube: https://www.youtube.com/@thinkingonpaper/videos

--

Coleman Collins, Director of Product at IonQ, joins Thinking on Paper to explain how trapped-ion quantum computers work and why IonQ believes they offer a credible path towards useful quantum computing.

IonQ builds its systems using individual atoms held in place and controlled with lasers. This differs from the superconducting, neutral-atom, photonic and topological approaches being pursued by companies including IBM, Google, Microsoft and D-Wave.

In this episode, Coleman discusses how these competing quantum-computing architectures should be evaluated and why headline qubit counts reveal little about a machine’s practical performance.

We cover:

• How trapped-ion quantum computing works

• Why fidelity, connectivity and control matter more than raw qubit counts

• What IonQ means by “algorithmic qubits”

• How lasers are used to control and connect trapped-ion qubits

• The technical criteria required for a viable quantum computer

• When quantum computers could deliver a practical advantage

• Whether quantum computing could eventually threaten modern cryptography

• How IonQ’s approach compares with superconducting and topological quantum computing

• What investors misunderstand about quantum-computing performance

• How developers can begin working with quantum systems today

We also examine Microsoft’s Majorana-based strategy, the limitations of current quantum hardware and the difference between measurable technical progress and corporate claims.

This is a practical discussion about IonQ, trapped-ion quantum computing and the race to build commercially useful quantum computers.

Please enjoy the show.

--

Chapters:

(00:00) Introduction to Quantum Computing and IonQ

(02:17) Understanding Trapped Ion Quantum Computing

(04:57) DiVincenzo's 5-Step Criteria for Quantum Computers

(07:31) The Natural Aspect of Quantum Computing

(10:01) Algorithmic Qubits vs Physical Qubits

(12:49) Achieving Quantum Advantage

(15:04) Investment Trends in Quantum Computing

(17:44) The Role of Education in Quantum Investment

(20:40) Hot Buttons

(25:57) Topological Quantum

(30:47) Microsoft's Majorana Quantum Chip

(34:09) Engaging Developers in Quantum

(36:59) Hybrid Quantum Computing and Integration

(38:55) The Quantum Promised Land

(41:19) Can Quantum Hack Bitcoin?

(45:09) Could Quantum Currency Exist?

--

www.thinkingonpaper.xyz

Philip Johnston, co-founder of StarCloud, joins Mark and Jeremy to explain how the company plans to build AI data centers in space.

StarCloud is developing orbital computing infrastructure designed to run large numbers of GPUs using solar power. Instead of relying on water-based cooling systems, the company plans to remove heat through large space radiators and transmit data between orbit and Earth using optical links.

The economics of the idea depend heavily on falling launch costs. As reusable rockets reduce the cost of reaching low Earth orbit, companies such as StarCloud are exploring whether energy-intensive computing could eventually be moved away from terrestrial power grids and data centers.

In this episode, we discuss:

• How data centers in space would work

• Why StarCloud wants to place AI GPUs in low Earth orbit

• How space-based computers can be cooled without water

• Whether solar power could reduce the energy costs of AI computing

• How laser communication links compare with terrestrial fiber

• Where orbital networks could offer lower latency

• How falling SpaceX launch costs affect the economics of space infrastructure

• What hardware StarCloud plans to launch

• The engineering and commercial challenges facing orbital data centers

Philip also explains why StarCloud believes space-based computing could support the growing demand for AI infrastructure without placing the same pressure on terrestrial electricity grids and water supplies.

The conversation examines whether orbital computing is a practical extension of cloud infrastructure or an ambitious response to problems that are still cheaper to solve on Earth.

Please enjoy the show and share it with a curious friend.

--

Chapters

(00:00) The impact of earth based compute

(03:09) Data centers in space

(08:36) Conquering Latency Challenges in Space

(10:32) Modular Space Infrastructure

(16:03) How much do space based data centers cost?

(19:46) Manufacturing Beyond Earth's Boundaries

(26:00) Reusability and space junk

(26:15) GPUs in Orbit

(28:52) Future Tech Rapid-Fire Questions

(29:55) 5 Billion Humanoids

(32:35) Addressing Space Skepticism

(33:41) Quantum Computing's Orbital Advantage

The Humanoid Difference

(37:20) Where Are All the Aliens? Exploring the Fermi Paradox

(38:35) Behind the Scenes

--

Other ways to connect with us:

• ⁠Listen to every podcast⁠

• Follow us on ⁠Instagram⁠

• Follow us on ⁠X⁠

• Follow Mark on ⁠LinkedIn⁠

• Follow Jeremy on ⁠LinkedIn⁠

• Read our ⁠Substack⁠

• Email: [email protected]

In Chapter 9 of Nexus, Yuval Noah Harari examines whether democracy can survive the data age and what happens when artificial intelligence begins to compete with humans in work, creativity and decision-making.

Mark and Jeremy discuss Harari’s argument that AI could disrupt intellectual professions before many manual jobs. They also examine whether the deeper risk is not simply unemployment, but the loss of human agency, economic independence and purpose.

The episode explores four principles Harari proposes for democratic societies adapting to AI: benevolence, decentralisation, mutuality and rest. These principles are intended to limit concentrated power, distribute control and give institutions time to understand technologies before adopting them at scale.

In this episode, we discuss:

• Whether democracy can survive increasingly powerful AI systems

• Why intellectual and administrative jobs may be automated before manual work

• What Harari means by benevolence, decentralisation, mutuality and rest

• Whether human creativity can remain distinct from machine-generated output

• What children should learn in a labour market shaped by AI

• How free digital services exchange convenience for personal data

• Whether individuals should own or control the data generated about them

• How AI could affect human agency, meaning and political participation

This isn’t only a discussion about whether AI will replace jobs. It’s about who controls the systems that make decisions, how democratic societies can hold those systems accountable, and what remains uniquely human when machines can produce language, images and ideas at scale.

--

TIMESTAMPS:

(00:00) Tech optimism(

01:29) Chapter 9 first impressions

(03:14) The race for raw materials

(06:02) Benevolence, decentralization, mutuality, change & rest

(15:30) Automation assumptions are wrong

(18:18) Motor skills and social skills in 2050

(20:06) The flexible superpower

(23:11) Creativity: the last refuge of the human condition

(26:53) Does AI have exceptional test?

(28:50) Reverse creativity

(31:45) Go "Move 37"

(33:52) Bank algorithm says no

(37:15) Your AI girlfriend

--

Learn more: https://www.thinkingonpaper.xyz/technology-book-club/

Fintan Buckley, CEO of Ubotica, joins Thinking on Paper to explain how artificial intelligence is changing the way satellites observe Earth, detect disasters and make decisions in orbit.

Traditional Earth-observation satellites often collect data and send it back to ground stations for processing. Ubotica is developing onboard AI systems that can analyse images in space, identify relevant events and transmit useful information without waiting for the full dataset to reach Earth.

This could reduce the time needed to detect wildfires, monitor environmental changes and respond to maritime incidents. It also raises questions about surveillance, military use and how much decision-making should be delegated to autonomous systems.

In this episode, we discuss:

• How AI-powered satellites work

• How satellites can detect wildfires in near real time

• Why processing data in orbit can reduce disaster-response times

• How satellite AI supports maritime security and environmental monitoring

• The privacy implications of increasingly precise Earth observation

• How autonomous satellites could be used in military operations

• Whether humans should remain involved in critical decisions

• Why removing old satellites won’t solve the wider space-debris problem

• How realistic fictional surveillance satellites are

Fintan explains where onboard satellite computing is already being used, what technical limitations remain and how close the industry is to deploying more autonomous systems in orbit.

The conversation examines both sides of AI-powered Earth observation: its potential to improve disaster response and environmental monitoring, and the risks created when satellites can analyse, prioritise and act on information with limited human involvement.

Please enjoy the show.

--

📌 Stay Connected:

🌍 Website: ⁠ThinkingOnPaper.xyz⁠ 

📩 Email: ⁠[email protected]⁠ 

🎧 Podcast: Available on all platforms

🔔 Subscribe for sharp, no-fluff insights on AI, space, and the future of technology.

In Chapter 8 of Nexus, Yuval Noah Harari examines the AI alignment problem: how to ensure that artificial intelligence acts in ways that support human goals without causing harmful or unintended consequences.

The central risk isn’t necessarily that AI will rebel. A system can follow its instructions accurately and still produce disastrous results because its objective is incomplete, poorly defined or detached from the context in which it operates.

Mark and Jeremy discuss Harari’s examples of obedience, incentives and unintended outcomes, from Stalin-era loyalty tests to modern recommendation algorithms.

In this episode, we discuss:

• What the AI alignment problem is

• Why obedient AI systems can still cause harm

• How Stalin’s applause test illustrates dangerous incentive structures

• What Napoleon’s victories reveal about intelligence and long-term judgement

• How the paperclip maximiser thought experiment explains misaligned objectives

• Why fixed rules such as Asimov’s Three Laws can’t resolve every ethical conflict

• How social media algorithms manipulate attention and emotion

• Why current recommendation systems already demonstrate alignment failures

• Whether AI safety can be reduced to rules, constraints or technical safeguards

The episode distinguishes between rogue AI and a more immediate problem: systems that pursue the goals humans give them without understanding the values, trade-offs and consequences behind those goals.

AI alignment isn’t only about stopping machines from disobeying us. It’s about deciding what we should ask them to do, how those objectives should be interpreted and who bears responsibility when the result causes harm.

Please enjoy the show.

--

Timestamps

[00:00] Introduction: Books That Change Minds

[01:04] Diving into Nexus Chapter 8

[01:37] The Stalin Test: When Applause Becomes Terror

[06:11] Evolution of AI Principles

[07:45] Understanding the Attention Economy

[08:45] How AI Targets Our Limbic System

[09:29] Inside Facebook: The Leaked Reports

[11:49] Napoleon's Warning for AI[

15:55] The AI Alignment Problem Explained

[17:49] Racing Against Time: Human Goals vs. Doomsday Clock

[20:04] The Power of Divergent Thinking

[21:50] Understanding Deontology in AI Ethics

[26:55] Can Mythology Guide AI?

[27:54] Exploring Inter-computer Realities

[33:50] Why Asimov's Laws Won't Save Us

[38:31] NPCs & The Future of Digital Consciousness

Boris Mamlyuk and Anuj Dasgupta, founders of STXN, join Thinking on Paper to explain how smart transactions could make Ethereum easier to use.

Standard blockchain transactions execute immediately under fixed conditions. Users often have limited control over when a transaction runs, what fees they’ll pay or what happens if market conditions change before execution. Failed transactions can still consume gas, leaving users with a fee but no completed action.

STXN is developing programmable transactions that allow users and applications to define how, when and under what conditions blockchain activity should occur.

In this episode, we discuss:

• What smart transactions are and how they work

• Why Ethereum transactions fail

• Why failed transactions can still consume gas

• How users could control transaction timing, fees and execution conditions

• How scheduled cryptocurrency payments could work

• What a decentralised time oracle is

• Why blockchains need reliable ways to coordinate time

• How conditional execution could improve Ethereum applications

• What programmable transparency means for companies and governments

• Whether better transaction infrastructure can make blockchain useful to mainstream users

Boris and Anuj also explain why existing crypto products remain difficult for ordinary users and how programmable execution could support payments, decentralised finance and automated blockchain applications.

The conversation examines whether smart transactions can reduce failed execution, wasted gas fees and the complexity that continues to limit Ethereum adoption.

--

🎧 Listen now and stay ahead of the next wave of blockchain innovation.

🔗 Get the STXN app & learn more: STXN.io

🔗 Follow Us: www.thinkingonpaper.xyz

--

Timestamps:

(00:00) Ethereum’s biggest problem?
(04:30) Why Ethereum transactions fail
(08:00) What are smart transactions
(14:10) How Ethereum’s UX is broken and what STXN is doing about it
(17:45) The traffic jam problem, or why Ethereum gets clogged
(23:34) Decentralized time oracle
(36:02) Google, AI & surveillance
(42:33) The crypto philosophy
50:10) Blockchain incentives
(55:57) Time-locked transactions

--

#ethereum #blockchain