In Focus

About In Focus

In Focus is your go-to podcast for exploring the trends, topics, and events shaping our world today. From technology and culture to global news and lifestyle, we bring you what’s in focus and what everyone should know about. Join us as we break down the stories that matter, offering fresh insights and engaging conversations to keep you informed and inspired. Stay curious, stay connected—stay In Focus.

Publisher

In Focus

Category

technology

Language

en

Explicit

No

First episode

28 Dec 2024

Latest episode

4 Oct 2026

Latest episodes

11 episodes in the feed.

1. 

   4 Oct 2026

   When AI Agents Go Rogue: Who Is Responsible?

   AI agents are no longer just answering questions. They can browse the internet, write and deploy code, access databases, use credentials, make decisions and take actions on behalf of humans. But what happens when an AI agent goes rogue? If an autonomous agent deletes production data, makes an unauthorized payment, leaks confidential information, attacks another system, or signs a contract it was never supposed to sign, who is legally responsible? The developer who built the model? The company that deployed it? The employee who gave it access? The vendor that supplied the agent? Or nobody? In this episode, we investigate the emerging legal and cybersecurity accountability problem created by agentic AI. We examine real-world incidents, existing liability principles, AI agent identity and authorization, negligence, product liability, corporate responsibility, human oversight and the uncomfortable gap between an AI being technically autonomous and legally responsible. Because an AI agent may be able to act on its own. But legally, it still can't stand in court and take the blame. So when the machine goes rogue… Who pays?

2. 

   27 Sept 2026

   The Race to Control AI Before AI Controls Us

   For the first time, the people building some of the world's most powerful AI systems are taking the AI safety debate to the United Nations Security Council. Sam Altman of OpenAI and Dario Amodei of Anthropic warned world leaders about increasingly autonomous AI, misuse, cybersecurity, and the possibility that future systems could become difficult for humans to control. But there is a deeper question hiding behind the headlines: Why are the companies racing to build more powerful AI now asking governments to help control it? In this episode, we go beyond the UN speeches and examine the real conflict underneath them — innovation versus control, national competition versus international cooperation, economic opportunity versus systemic risk, and voluntary safety commitments versus enforceable standards. Are we genuinely approaching a point where AI could escape meaningful human control? Or are today's warnings being shaped by geopolitical competition, commercial incentives and uncertainty about the future? And if AI becomes a strategic technology comparable to nuclear weapons, cybersecurity and critical infrastructure, who should ultimately decide how powerful it is allowed to become? The AI debate has moved from Silicon Valley to the world's most powerful diplomatic table. The question is no longer simply what AI can do. It is who gets to decide what AI is allowed to do.

3. 

   22 Sept 2026

   Jev: The AI That Doesn’t Talk - It Decides

   What if the next big evolution in AI isn’t a smarter chatbot? Jev is taking a radically different approach: instead of generating text, it makes fast, structured decisions that software can act on. TypeSafe AI calls this a “System One Model” — an AI designed not to talk to humans, but to become a decision-making layer inside software and autonomous agents. In this episode of In Focus, we explore why this seemingly small change could have enormous consequences. Could AI become dramatically cheaper and faster when we stop asking it to explain everything? What happens when millions of software decisions that once required rules, humans, or expensive LLM calls can instead be handled by a tiny AI decision engine? And could this be the missing infrastructure for truly autonomous agents? We examine Jev, System One models, calibrated decisions, AI agent architectures, automation, hallucination and reliability — and the bigger question: are we moving from AI that talks to us toward AI that quietly runs the software around us?

4. 

   17 Sept 2026

   Should We Slow AI Down? The Race Between Progress and Catastrophe

   AI is advancing faster than our ability to understand, control, and regulate it. But should we actually slow it down? In this episode of In Focus, we examine the growing debate over “pacing” AI development: the argument that frontier AI needs to slow down long enough for safety, security, and governance to catch up. Are existential AI risks genuinely credible—or are we letting science-fiction scenarios distract us from the very real economic and social disruption already happening? Could slowing AI development reduce catastrophic risks, or would it simply hand technological and geopolitical influence to whoever keeps building? We put the strongest arguments on both sides under the microscope: economic growth versus inequality, innovation versus safety, autonomous agents versus human control, and AI leadership versus global coordination. No hype. No doomism. Just one difficult question: how fast should humanity build a technology that may eventually become more capable than its creators?

5. 

   16 Sept 2026

   The Quantum Countdown: Is the Internet Ready for the Quantum Threat?

   Description What happens when a quantum computer becomes powerful enough to break the cryptography protecting today’s internet? In this episode of In Focus, we explore the coming post-quantum cryptography transition — from NIST’s finalized quantum-resistant standards to Cloudflare’s real-world deployment of post-quantum DNSSEC on its 1.1.1.1 resolver. The story is bigger than encryption. DNS is one of the internet’s foundational systems, and Cloudflare’s experiment shows just how difficult it can be to upgrade global infrastructure without breaking compatibility, performance, or trust. Its new DNSSEC implementation uses ML-DSA-44, a post-quantum signature whose 2,420-byte size creates very practical networking challenges. We also look at NIST’s three finalized standards — ML-KEM, ML-DSA and SLH-DSA — and why organizations are being urged to begin migration now, even though large-scale quantum computers capable of breaking today’s public-key cryptography do not yet exist. From “harvest now, decrypt later” to the possibility of a future where a compromised quantum-era key could undermine trust across the DNS hierarchy, this episode asks a simple question: Are we preparing for the quantum threat early enough — or will the internet discover the problem only when it is too late?

6. 

   15 Mar 2025

   The Enshittification Chronicles: How Big Tech Killed Innovation (And Why China is Winning)

   Join us on "The Enshittification Chronicles" as we uncover how U.S. tech giants once hailed as innovators have increasingly sacrificed genuine creativity for short-term profit—transforming platforms like Facebook into echo chambers of ads and propaganda. We explore why these stagnating trends are creating opportunities for bold, risk-taking Chinese companies, and what this shift means for the future of digital innovation.

7. 

   24 Feb 2025

   AI Meets Matter: The Story of MatterGen's Pioneering Material Discoveries

   In this episode of "In Focus," hosts Emily and Michael explore MatterGen, Microsoft's groundbreaking AI tool that's transforming materials science. Discover how MatterGen accelerates the discovery of novel materials by generating stable, functional compounds tailored to specific needs, a process that traditionally took years. Learn about the history of materials discovery, the challenges scientists faced, and how MatterGen's diffusion modeling technique is revolutionizing the field. We'll also discuss the future implications of this technology, from advancements in electronics to renewable energy solutions, and how the open-source nature of MatterGen fosters global collaboration. Join us as we delve into the dawn of a new era in advanced technology, where AI and human ingenuity converge to propel scientific discovery forward.

8. 

   28 Jan 2025

   Open-Source vs. OpenAI: Is DeepSeek Rewriting the Rules of the AI Arms Race?

   Welcome to our thought-provoking podcast episode, "Open-Source vs. OpenAI: Is DeepSeek Rewriting the Rules of the AI Arms Race?" In this riveting discussion, we delve into the groundbreaking emergence of DeepSeek, a Chinese AI startup that's challenging the status quo of artificial intelligence development. We explore how DeepSeek's open-source model, DeepSeek-R1, is not only matching the performance of industry giants like OpenAI but doing so at a fraction of the cost. Our expert panel examines the implications of this technological leap, from its potential to democratize AI development to its impact on data privacy and innovation scalability. We'll dissect the ongoing debate between open-source and proprietary AI models, considering factors such as cost-effectiveness, customization capabilities, and long-term strategic value for businesses. Join us as we unravel the complexities of this AI revolution and its far-reaching consequences for the tech industry, global competition, and the future of artificial intelligence. Whether you're a tech enthusiast, business leader, or simply curious about the cutting edge of AI, this episode promises to deliver insights that will reshape your understanding of the AI landscape.

9. 

   24 Jan 2025

   LA Wildfires: Why it is a big deal ?

   The overall estimated financial losses from the January 2025 Los Angeles wildfires are substantial and are projected to be among the costliest in U.S. history . Here's a breakdown of the estimated losses: Total Estimated Losses: Preliminary estimates range between $135 billion and $150 billion . Some sources indicate losses between $52 billion and $57 billion . These losses are due to the destruction of properties, infrastructure, and the disruption of business activities. Insured Losses: Insured losses are projected to exceed $20 billion, potentially setting a new record for wildfire-related insurance claims in U.S. history . Housing Market Impact: The destruction of numerous homes has caused a surge in rental prices, with some landlords increasing rents by 15-20%, further exacerbating the financial strain on displaced residents . The fires are expected to worsen California's already stressed insurance market as claims increase . Economic Repercussions: The fires are on track to be among the costliest ever seen in the US . Property Damage: Over 12,401 structures have been damaged or destroyed . The Palisades fire alone is responsible for about 5,000 and the Eaton fire for 7,000 . The significant financial impact of these wildfires underscores their severity and the challenges in recovery and rebuilding .

10. 

    6 Jan 2025

    Navigating the Future of Cyber Defense- Agentic AI, Automation, and the Future

    Join us as we explore the dynamic evolution of Security Operations (SecOps). From manual processes to the rise of AI and automation, we delve into the key phases of SecOps maturity recently published by SentinelOne, the shift from SOAR to intelligent agents, and the crucial role humans will play in the cyber landscape of tomorrow. Reference: https://www.sentinelone.com/blog/introducing-the-autonomous-soc-maturity-model/

11. 

    28 Dec 2024

    The State of Cybersecurity in 2024: A Full-Year Report

    Key Themes and Trends: Ransomware Remains a Dominant Threat: High financial impact and operational disruption continued, exemplified by attacks on UnitedHealth Group and LockBit's resilience. Sophistication of Nation-State Actors: Espionage and potential infrastructure disruption remained significant, with the Microsoft email breach as a key example. Supply Chain as a Vulnerable Entry Point: Exploitation of MOVEit and Barracuda highlighted the risk from third-party software and services. Critical Infrastructure Under Siege: Attacks on essential services like water utilities demonstrated ongoing vulnerabilities. Exploitation of Known Vulnerabilities Persists: Unpatched software, as seen with TP-Link routers and Follina, remained a problem. Data Exfiltration as a Key Goal: Stealing sensitive data was a primary objective in many attacks, particularly ransomware. The Rise of AI in Attacks and Defense: Offensive AI: Used for personalized phishing, automated vulnerability discovery, deepfakes, and polymorphic malware. Defensive AI: Employed for advanced threat detection, behavioral analysis, automated incident response, and vulnerability management. Cloud Security Challenges: Misconfigured cloud environments continued to be a major cause of breaches. Zero-Day Exploits as a Continuing Threat: Barracuda and MOVEit attacks demonstrated the danger of previously unknown vulnerabilities. Growing API Security Concerns: The increasing use of APIs made them a significant attack vector. Top 10 Threats/Attacks in 2024: The report identified the following as the top ten threats throughout the year: Ransomware Attacks Exploitation of Known Vulnerabilities Supply Chain Attacks Nation-State Sponsored Attacks AI-Enhanced Phishing and Social Engineering Cloud Security Misconfigurations Zero-Day Vulnerability Exploitation Attacks on APIs Insider Threats Cyberattacks Targeting Critical Infrastructure

Reach and audience

Public platform figures. Ratings count people who left a rating, not total listeners.

Spotify

5.0 / 5

1 ratings