Tech Talks Daily

What happens when AI makes employees more productive today but gradually weakens the expertise companies will depend on tomorrow?
In this episode of Tech Talks Daily, I speak with Dr. Margaret Cunningham, VP of Security and AI Strategy and Field CISO at Darktrace, about cognitive tech debt, the growing risk that companies are gaining short-term efficiency from AI while unintentionally weakening critical thinking, technical expertise, problem-solving ability, and human judgment.
Margaret brings a rare combination of experience to this conversation. With a PhD in Applied Experimental Psychology and a career spanning behavioral science, cybersecurity, privacy, human-centered security, and AI strategy, she examines technology adoption through the lens of how people actually think, learn, develop expertise, and make decisions.
She explains cognitive tech debt by comparing it with the technical debt familiar to software teams. Companies can introduce technology quickly and enjoy immediate improvements in speed and output, only to discover weaknesses underneath those gains later. With AI, the debt may accumulate in people. Employees can appear highly productive while outsourcing the difficult cognitive work required to build judgment, recognize patterns, understand failures, and develop genuine expertise.
We discuss emerging evidence that over-reliance on AI is already affecting professional skills. Software engineers may become less capable of diagnosing problems in code they did not create themselves. Medical professionals can lose decision-making capabilities when they become dependent on automated systems. Across knowledge work, deep reading and sustained concentration are increasingly being replaced by summarization, generation, and superficial review.
Margaret describes the current period as the "bridge years," when AI systems are becoming increasingly capable but people still need to maintain the expertise required to recognize mistakes, question recommendations, recover from failures, and understand when automation should not be trusted. Companies cannot safely abandon human skills before technology can reliably perform those responsibilities without supervision.
The conversation also challenges one of the most repeated promises surrounding enterprise AI adoption: that automation will remove routine work and allow employees to concentrate on higher-value activities. Margaret argues that companies have done a poor job of defining which tasks people genuinely want to give up and which skills they need to preserve. Some of the repetitive, slow, and difficult work being automated may be exactly where people develop pattern recognition, creativity, and professional judgment.
This creates a serious challenge for cybersecurity teams and other high-stakes professions. If employees become reviewers of AI-generated outputs rather than practitioners developing expertise through experience, where will the next generation of senior engineers, security analysts, doctors, researchers, and technical specialists come from?
Margaret explains why leaders need to understand which AI techniques are being used for different business problems rather than treating every form of artificial intelligence as interchangeable. Large language models, machine learning systems, behavioral analytics, and other technologies have different strengths and limitations. Knowing what questions to ask requires domain expertise, creating a difficult paradox for companies that may be automating away the very experience needed to govern these systems responsibly.
We also examine the human consequences of AI adoption. Technical specialists who enjoy solving difficult problems can lose motivation when meaningful work is replaced by reviewing machine-generated outputs. Companies may struggle to understand who owns decisions made through collaboration between humans and AI, while younger employees could lose access to the experiences that previously helped people progress from beginners to experts.
Margaret offers practical advice for business and technology leaders deciding how quickly to introduce AI across their workforce. Companies can identify the skills they need to preserve, create opportunities for employees to practice difficult cognitive work, use simulations and training to maintain expertise, ask teams which aspects of their jobs give them purpose, and resist pressure to automate every task simply because the technology exists.
The message is not anti-AI. Margaret sees enormous potential for artificial intelligence in scientific research, cybersecurity, productivity, and solving difficult problems. But realizing those benefits requires a more intentional relationship between people and machines.
For business leaders, CISOs, technology teams, AI practitioners, and anyone concerned about the future of human expertise, this conversation provides a practical framework for recognizing cognitive tech debt, deciding what should and should not be automated, preserving critical thinking skills, and building healthier forms of human-AI collaboration.
AI can make people faster. The bigger question is whether companies can capture those productivity gains without losing the human capabilities they will need when the technology gets something wrong.

What happens to blockchain networks, digital assets, and the wider internet when quantum computers become powerful enough to break the cryptography protecting them?
In this episode of Tech Talks Daily, I speak with Bruno Martins, Chief Technology Officer of the Algorand Foundation, about what quantum computing means for blockchain security, why post-quantum cryptography is becoming a technology priority, and how enterprises should evaluate blockchain infrastructure for payments, digital assets, identity, and other business applications.
Bruno brings experience from across several major blockchain ecosystems, including Consensys and IOHK, alongside a background in applied cryptography, key management systems, enterprise blockchain development, and software engineering. His perspective provides a useful view of how the blockchain industry has changed from experimental projects and speculative use cases toward platforms expected to support real financial transactions and business operations.
We begin with the quantum threat itself. Bruno explains why the cryptographic systems protecting blockchains, financial infrastructure, communications, messaging platforms, and much of the internet could eventually become vulnerable to sufficiently powerful quantum computers. 
While the exact timeline remains uncertain, he argues that waiting for a cryptographically relevant quantum computer to arrive before beginning migration would leave companies with too little time to update infrastructure, applications, wallets, accounts, and user behavior.
The conversation examines why post-quantum security is not simply a future technology problem. Large digital ecosystems can take months or years to migrate, and businesses need time to understand their cryptographic dependencies, introduce new standards, educate users, and build systems capable of adopting new security methods without disrupting existing operations.
Bruno shares how Algorand has been working on post-quantum security for several years, including the deployment of Falcon signatures for state proofs and plans to introduce quantum-resistant account types and additional protections across consensus and network communications. We discuss why cryptographic agility may be more important than simply replacing existing cryptography with newer algorithms that have not yet experienced decades of testing in real-world systems.
This leads to one of the most valuable technical lessons in the episode. Moving directly from classical cryptography to post-quantum cryptography introduces its own risks because newer cryptographic methods may later reveal weaknesses. Bruno explains why hybrid approaches, where digital assets and accounts can be protected by both established and quantum-resistant cryptography, could provide a more responsible path for institutions managing long-lived systems and valuable assets.
We also examine how enterprises should evaluate blockchain platforms. With thousands of networks competing for developers, users, and institutional adoption, Bruno argues that businesses need to look beyond market attention and transaction speed. Throughput, decentralization, security, programmability, finality, operational risk, and the ability to trust the state of a ledger all influence whether blockchain infrastructure is suitable for real business operations.
Payments provide a practical example. Companies issuing payment products backed by stablecoins need confidence that transactions are final and cannot later be reorganized or reversed by the underlying network. Bruno explains why instant finality can reduce operational uncertainty and risk for companies building financial applications on public blockchain infrastructure.
The conversation also turns to AI agents and agentic commerce. If autonomous software agents begin negotiating, purchasing services, exchanging value, and conducting transactions with other agents, they will need payment rails, identity systems, trusted counterparties, and ways to establish ownership and accountability. Bruno explains why stablecoins, digital identity, decentralized finance, and blockchain infrastructure could become increasingly relevant as AI systems begin participating directly in economic activity.
Throughout the episode, Bruno offers a balanced assessment of the blockchain industry itself. He discusses the problems created by technical fragmentation, competing standards, thousands of networks, and ecosystem tribalism. Greater cooperation between blockchain communities, particularly around wallets, hardware, cryptographic standards, and post-quantum security, could make it easier for enterprises and developers to build applications that work across ecosystems.
For technology leaders, security professionals, blockchain developers, and anyone responsible for long-lived digital infrastructure, this conversation provides a practical introduction to quantum threats, post-quantum cryptography, cryptographic agility, blockchain finality, stablecoins, and the technical questions companies should ask before choosing distributed infrastructure.
The quantum threat may not arrive tomorrow, but migrating complex systems takes time. The companies and technology platforms preparing today will be in a much stronger position to protect digital assets, maintain trust, and continue operating when current cryptographic standards eventually need to change.

What if companies rushing to deploy AI agents are overlooking the basic problem that much of their business data is still trapped inside PDFs, emails, attachments, spreadsheets, and paper documents?
In this episode of Tech Talks Daily, I speak with Sylvestre Dupont, co-founder and CEO of Parseur, about why successful AI adoption begins with making business data usable, why traditional automation can often outperform more sophisticated AI systems, and how he built a profitable global technology company with six employees across six countries without venture capital funding.
Sylvestre introduces the concept of data liquidity, the ability to move information from the documents and systems where it is trapped into the applications, workflows, and AI systems that can put it to work. Companies may have years of valuable operational data, but if that information remains buried inside what Sylvestre calls "digital concrete," even the most advanced AI models will struggle to produce useful results.
The conversation examines why structured data extraction has become increasingly important as companies invest in AI agents, copilots, and automated workflows. Sylvestre explains that better models alone cannot compensate for incomplete, inaccessible, or poorly structured information. Before businesses can expect AI to automate complex processes or support better decisions, they need reliable ways to collect, structure, and move data between systems.
We also challenge the assumption that every business problem now requires an AI solution. Sylvestre explains why AI should be treated as one tool among many and why deterministic automation remains the better option for repetitive processes where accuracy, consistency, and explainability matter. Parseur itself combines AI-powered document processing with template-based extraction and traditional workflow automation, using each approach where it performs best.
Drawing on Parseur's experience processing more than 100 million documents annually, Sylvestre describes the different stages companies move through as they mature their automation strategies. Some begin by manually uploading documents and downloading extracted data. Others automate document ingestion and connect information directly to accounting platforms, CRM systems, and other business applications. The most advanced companies add exception handling and human review processes for situations where automation cannot reliably complete the task.
Data privacy and security are another major part of the discussion. Sylvestre shares the questions technology leaders should ask before sending sensitive company information to AI-powered platforms, including where data is stored and processed, whether customer information is used to train AI models, how deletion requests are handled, and whether vendors genuinely understand the regulations and security standards they claim to follow.
For founders and bootstrapped entrepreneurs, Sylvestre also shares an alternative perspective on building technology companies. Parseur has remained profitable, globally distributed, and customer-funded rather than pursuing the venture capital model of rapid expansion. Sylvestre explains why he prefers customers to determine the company's priorities, how asynchronous communication supports a team operating across multiple time zones, and why building a sustainable business can offer founders greater control over product decisions and company culture.
This conversation offers practical lessons for technology leaders deciding where AI belongs in their operations, operations teams trying to reduce repetitive manual work, and founders questioning whether venture capital is the only route to building a successful global software company.
The message throughout the episode is simple: AI can be extremely useful, but companies still need reliable data, appropriate technology choices, strong privacy practices, and well-designed business processes. Sometimes the smartest technology strategy begins by solving the boring problems first.

Why do companies continue spending heavily on cybersecurity technology while human behavior, poor governance, and skills shortages leave them exposed to attacks?
In this episode of Tech Talks Daily, I speak with Phil Chapman, Cybersecurity Subject Matter Expert at Firebrand Training, about what more than two decades in the Royal Air Force, signals intelligence, counterterrorism, threat intelligence, and cybersecurity education taught him about defending companies in an increasingly complex threat environment.
Phil's career provides a fascinating perspective on how intelligence skills developed in military and national security environments can be applied to modern cyber defense. After 23 years in the RAF, including work supporting organizations such as GCHQ and the NSA, training intelligence analysts, and working in counterterrorism, Phil moved into technology training and cybersecurity education. Today, he helps companies understand their cybersecurity training needs while supporting people building careers in an industry that continues to need new talent.
A major theme throughout our conversation is Phil's belief that cybersecurity is fundamentally about people. Technology matters, but expensive security products cannot compensate for employees who do not recognize threats, executives who misunderstand their responsibilities, or companies that treat security awareness as an annual compliance exercise.
Phil explains threat intelligence in practical business terms, examining the relationship between threats, vulnerabilities, business assets, and risk. We discuss why insiders remain one of the biggest security concerns facing companies, including malicious employees and the far more common problem of accidental actions such as clicking phishing links, sharing sensitive information, or sending data to the wrong recipient.
The arrival of generative AI is making these problems harder to manage. Phil discusses how criminals are using AI to create more convincing phishing campaigns, deepfakes, social engineering attacks, and other forms of cybercrime. At the same time, employees are introducing new risks by using AI tools without understanding what happens to company data or whether appropriate policies and controls are in place.
But this episode is also about opportunity. Phil challenges the stereotype that cybersecurity careers are only for highly technical people sitting behind multiple screens writing code. He explains the different career paths available across cybersecurity engineering, threat intelligence, incident response, security operations, governance, risk, compliance, and analysis, and why skills from customer service, the military, data analysis, writing, communications, and other professions can transfer successfully into cyber roles.
For anyone considering a career change or trying to enter the technology industry, Phil offers practical advice on where to begin. Rather than chasing advanced certifications or trying to become an ethical hacker immediately, he recommends building a strong foundation, understanding networks and operating systems, staying current with the news, developing analytical thinking, and remaining curious about how criminals adapt world events and new technologies to create attacks.
We also discuss cybersecurity apprenticeships and why alternative routes into technology careers could help companies develop talent while giving people of different ages and professional backgrounds access to an industry they may previously have considered out of reach.
Finally, Phil explains why cybersecurity professionals cannot focus only on today's threats. AI is already changing both attack and defense strategies, while quantum computing is forcing companies to examine cryptography, data protection, and long-term security planning. His message to business leaders and technology professionals is clear: buying more technology will not solve every security problem. Companies need informed leadership, better governance, continuous learning, practical training, and people who understand how threats evolve.
This conversation offers business leaders a clearer understanding of cyber risk, provides technology teams with practical ideas for improving security awareness, and offers anyone considering a cybersecurity career a realistic view of the opportunities, skills, and pathways available through training and apprenticeships.

What separates the founders who build category-defining companies from the thousands of startups that never make it through the venture capital funnel?
In this episode of Tech Talks Daily, I speak with Keval Desai, founder and General Partner of Shakti, an early-stage venture capital firm investing in AI and space technology companies from inception. Drawing on his experience backing companies including Canva, The RealReal, and Gatik, Keval shares how he evaluates founders before the rest of the market recognizes their potential and why the venture capital industry needs to confront some uncomfortable truths about startup funding and successful exits.
Keval introduces Shakti's "toothbrush" investment philosophy, an idea he first encountered through Larry Page at Google. The principle is simple: can a product or service become something used frequently by millions or even billions of people? He explains why this question helps investors distinguish impressive technology from businesses capable of creating lasting value, particularly at a time when thousands of AI startups are competing for capital and attention.
But identifying a large market is only part of the equation. Keval shares three characteristics he has observed in exceptional founders. They can describe a future that others cannot yet see, attract talented people before they have money or resources, and execute at a speed that continually surprises those around them. His stories from meeting Canva co-founder Melanie Perkins and The RealReal founder Julie Wainwright offer a rare look at what investors can learn from founders at the earliest stages of company building.
We also discuss Keval's thesis that AI is taking the economy into a new Imagination Era. As AI becomes increasingly capable of handling specialized tasks such as coding, analysis, and production, he believes human value will move toward imagination, judgment, taste, and the ability to combine technologies into products and services people actually want. For founders, employees, and business leaders, this raises important questions about education, careers, and what it means to build a company as access to technical capabilities becomes dramatically cheaper.
Keval also compares the arrival of open-source AI models such as DeepSeek to the role Linux played in the development of the commercial internet. He explains why falling inference costs could lower barriers to building AI companies and create opportunities for a new generation of startups, while also examining what this could mean for today's dominant AI companies and the industry's economics.
The conversation then turns to one of the biggest problems facing venture capital. The number of startups receiving funding has grown dramatically, yet the number of technology companies reaching public markets has remained relatively static. Keval explains why venture capital can scale dollars but cannot simply manufacture more category leaders, and why founders need to decide early whether venture capital is actually the right source of funding for the business they want to build.
We also examine the commercial opportunities emerging from space technology. Keval believes the SpaceX IPO could play a similar role for space commerce to Amazon's IPO for e-commerce, by demonstrating viable business models and encouraging entrepreneurs to build new companies in communications, energy, manufacturing, infrastructure, robotics, and services beyond Earth.
Finally, Keval offers an optimistic counterargument to fears that AI will leave younger workers without meaningful careers. He explains why he believes Gen Z's status as the first AI-native generation could become an advantage, why technical careers are changing rather than disappearing, and why the ability to apply AI to problems across healthcare, manufacturing, agriculture, finance, and other industries could create opportunities far beyond Silicon Valley.
This conversation offers founders a practical framework for evaluating ideas, choosing investors, understanding venture economics, and building companies in the age of AI. It also provides investors and technology leaders with a broader perspective on open-source AI, space commerce, the future of work, and where the next generation of category-defining companies could come from.