The Entropy Podcast

Former BBC foreign journalist Melanie Marshall spent more than 20 years reporting from conflict zones and moments of upheaval across six continents. In this episode, she reveals what those experiences taught her about leadership, resilience and making decisions when you don’t have the full picture.
Melanie and Francis explore why uncertainty does not mean powerlessness, how great leaders create space for uncomfortable truths, and why plans must change the moment they meet reality. They also examine AI’s growing influence on human judgment, the dangers of easy answers and why businesses must resist handing over their agency.
From a Ukrainian community that fought back against an invasion to a life-saving decision on a road in Syria, Melanie shares remarkable stories that demonstrate courage, trust and leadership under pressure.
Key Takeaways:
Uncertainty is not powerlessness it creates both risk and opportunity.
Effective leaders make decisions with incomplete information.
A plan must evolve when it comes into contact with reality.
Leadership means confronting uncomfortable truths before the window for action closes.
Trust must be built before a team reaches a crisis.
The quietest person in the room may recognize your greatest risk or opportunity.
AI should support human judgment, not replace curiosity, accountability or agency.
Resilience often begins with identifying the resources already available to you.
Human connection remains one of our strongest defenses against fear and division.
Courage is often as simple as taking the next step together.
Soundbites:
“Uncertainty is not powerlessness.”
“The plan survives usually until you leave the door.”
“You’re deciding with the map actively on fire.”
“Bravery is simply one foot in front of the other.”
“Leadership is doing the uncomfortable thing.”
“We need a little grit to make the pearl.”
“Go out into the world and talk to a real person.”
“If the world feels scary, ask yourself: What is the opportunity here, and what resources do I have available?”
Listen now for a compelling conversation about leading through uncertainty, decision-making under pressure, resilience, AI, human connection and the courage to move forward without all the answers.

If you haven’t done so already check out part 1.
Apple podcast: https://podcasts.apple.com/ie/podcast/the-entropy-podcast/id1795927497?i=1000788046671
Spotify: https://open.spotify.com/episode/2EiCIRM6HhicJco4c8ZQSm?si=_MfmoPrzTXe6TiRIgUU-XQ&utm_source=copy-link

In Part 2 of Francis Gorman’s conversation with technologist and entrepreneur Paul Walsh, the discussion moves from online identity into a much broader question: what happens to privacy when AI becomes embedded in the tools we use every day?
Paul argues that the risks are no longer confined to social media platforms. As AI assistants gain access to messaging, email, devices and personal data, he raises concerns about what that means for end-to-end encryption, confidential conversations and the expectation that private communication remains private.

From there, the conversation widens into anonymity, censorship, AI-driven decision-making and the increasingly blurred relationship between governments, regulators and major technology companies.

For Paul, the danger isn’t necessarily one dramatic moment where privacy disappears. It is the gradual accumulation of systems built for convenience, safety or security that ultimately create much greater capabilities for surveillance and control.

Key Takeaways

End-to-end encryption is only as strong as its endpoints. Paul argues that giving an AI assistant access to messages before encryption or after decryption can undermine the practical privacy users expect from encrypted communication.
Convenience comes with a privacy trade-off. AI tools can make it easier to search conversations, schedule actions and retrieve information but those capabilities may require extensive access to sensitive personal data.
Privacy doesn’t require justification. One of Paul’s central arguments is that people should not have to prove why they deserve privacy or anonymity.
Identity verification changes the consequences of being banned online. Paul warns that if every account becomes permanently linked to a government identity, losing access to a platform could become far more consequential.
The chilling effect may matter as much as censorship itself. If people believe their identity is permanently attached to everything they say, Paul argues they may begin censoring themselves before anyone else has to.

Soundbites

“If somebody can read the message before you send it and before it’s encrypted, or after somebody sends you a message and after it’s decrypted, then you defeat the purpose.”

“People make a decision and think it was their own idea.”

“We have people lobbying for child safety, not realizing that they’re actually begging the government to surveil them.”

“Convenience and security have always been at odds.”

Part 2 leaves listeners with a difficult question: how much privacy are we willing to exchange for convenience, safety and access to increasingly powerful technology?

Paul’s message isn’t to abandon AI or digital platforms. It is to understand the systems being built around us, question who benefits from them, and protect the ability to communicate, participate and speak without unnecessary surveillance.

Because, as Paul puts it:

“I don’t have to explain to you why I want privacy.”

In Part 1 of this two-part conversation, Francis Gorman sits down with technologist and entrepreneur Paul Walsh to explore a question that has followed the internet for nearly three decades: how do we know what and who to trust online?
Paul traces the problem back to the early web and his work on content classification and verification standards, explaining why today’s challenges aren’t necessarily new they’ve simply been amplified by AI, deepfakes, phishing and increasingly sophisticated social engineering.
The conversation then turns to the growing push for online age verification and social media restrictions. Paul draws a sharp distinction between content moderation and access control, arguing that modern devices already give parents powerful tools to manage what children can access.
But his bigger concern is what comes next: when “robust age verification” evolves into mandatory identity verification for entire populations.
Key Takeaways
The trust problem hasn’t changed it has scaled. Social engineering and impersonation remain at the heart of much cybercrime; AI simply makes them faster, cheaper and harder to detect.
Verification is as much about psychology as technology. A badge or trust signal only works if people understand what it actually represents.
Deepfakes raise the stakes. In a world where people can no longer automatically trust what they see or hear, reliable provenance and account verification become increasingly important.
Content moderation and access control are different problems. Paul argues that platforms should improve moderation, while parents already have increasingly sophisticated device-level controls available to manage children’s access.
Age verification can become identity verification. Paul warns that demands for “robust” age checking could ultimately require users to prove their real-world identity simply to participate online.
Privacy can be lost incrementally. Measures introduced for legitimate goals such as child safety can expand in scope and create infrastructure capable of much broader surveillance.
Soundbites
“Society doesn’t know where to look to find trust.”
“Technology is just an enabler.”
“I don’t think any tech company is doing enough to reduce the risk of their technology being used for harm.”
“Content moderation is different to access control.”
“The governments are saying they’re giving back control, but actually they’re taking away their ability to just be a parent.”
Part 1 ends with a fundamental tension: how do we make the internet safer without building systems that require everyone to surrender their identity and privacy in order to participate?

AI is moving beyond automation and into autonomy and that shift could fundamentally change how businesses operate, how decisions are made and how people work.
In this episode of The Entropy Podcast, Francis Gorman sits down with Ismail Amla to explore the rise of agentic AI, digital workers and AI-native organisations. They discuss why the biggest opportunity is not simply using AI to make existing processes faster, but redesigning businesses around what intelligent systems can now do.
They also examine the harder questions: what happens when AI agents begin making decisions independently, who is accountable when those decisions go wrong, how organisations can avoid agentic drift, and whether relying too heavily on AI risks weakening human judgement and curiosity.
Key takeaways: 
Why AI is shifting from automation to autonomy
How AI agents could become part of the workforce 
Why most AI transformation efforts fall short
What leaders need to rethink, and why governance and human judgement matter more as systems become increasingly autonomous.
Soundbites: 
“I think it is a revolution. I don’t think it’s just another tool.”
“Whoever this technology is working on behalf of is responsible for every decision that’s made.”

In this episode of The Entropy Podcast, Francis Gorman sits down with Snehal Antani, CEO and co-founder of Horizon3.ai for a wide-ranging conversation on cybersecurity, AI, warfare, leadership and the future of work.
Snehal shares lessons from building and scaling Horizon3.ai, why startups must eventually move from “pirates” to a “navy,” and how his time inside Special Operations fundamentally changed his approach to leadership.
They also explore the rapidly changing cyber threat landscape: AI-powered attackers, autonomous penetration testing, deception as a defence against AI agents, the lessons emerging from Ukraine, and why organisations need to stop asking whether they are secure and start proving they are resilient and defensible.
Key Takeaways
Why great founders need grit, conviction and a “learn-it-all” mentality
How companies transition from entrepreneurial pirates to a scalable navy
Why AI could give cyber attackers effectively unlimited ammunition
How defenders can use honeypots and deception to hack the hackers
Why cybersecurity teams should train like they fight
The growing importance of human-machine teaming in warfare
Why over-reliance on AI could undermine human judgement and scepticism
How AI may reshape the workforce and put increasing pressure on the “middle”
Why there is no AI easy button coming for cybersecurity
Why organisations must move from being “secure” to being resilient and defensible
Key Soundbites
“You earn the right to be on this team every single day.”
“AI gives the adversary unlimited bullets.”
“You don’t want to learn how to deal with a crisis in the middle of a crisis.”
“We always train like we fight.”
“There is no AI savior that’s going to suddenly make your lives better.”
“Those obsessed with mastering their craft are going to come out on top.”
“Stop saying that we’re secure… start talking about how we’re defensible.”