Artificial intelligence is no longer a distant concept or a laboratory experiment. It is embedded in our interfaces, shaping how we think, decide and act.

In this episode of BASELINE, I sit down with Esra Kasapoglu to explore what happens when knowledge becomes abundant, agency becomes distributed, and AI moves from being a tool to becoming what she calls a “cognitive prosthetic.”We discuss how generative AI is reshaping power structures inside corporations, governments and everyday life.

If everyone now has access to knowledge, where does real power sit? Who designs the incentives that shape behaviour? And what does it actually mean for a country to become an “AI superpower”?

https://tig.foundation

In this episode of BASELINE, Ian speaks with Ying Chan, CTO of The Innovation Game. As AI accelerates software development, much of the conversation focuses on speed.

But speed does not make systems resilient. Incentives do. The Innovation Game is a protocol designed to incentivise the development of state-of-the-art algorithms through open participation. Innovators submit code. Benchmarkers evaluate performance. Incentives shape behaviour.

The strongest solutions emerge. Ying explains why AI-generated code is a powerful tool for first-pass execution, but not a substitute for strategic thinking or adversarial design. While models can interpolate from existing knowledge, pushing the frontier still requires human judgment, intuition, and carefully aligned incentives.

3/3 - If AI systems are going to be used in the real world, they have to work. This final film in the UK AI Reality series focuses on where generative AI projects actually succeed or fail, and why trust, quality, and control matter more than capability or scale. Jack Perschke speaks from direct experience about delivery, governance, auditability, and the practical constraints that shape whether AI systems can be relied on in public services and critical infrastructure. Rather than debating futures or hypotheticals, the conversation stays grounded in how systems behave once they are deployed, and what it takes to make them dependable over time. This is where the series closes, not with opinion, but with operational reality.

2/3 - Most conversations about AI assume someone is in control. This one questions that assumption. In this second film in the UK AI Reality series, the discussion moves beyond the founder and into the systems that shape national infrastructure, public services, and technological risk. Jack Perschke speaks candidly about how decisions around cloud, platforms, and AI are often responses to circumstance rather than deliberate strategy, and what it means when core infrastructure is dominated by a small number of global players.

The conversation ranges from sovereignty and public sector delivery to governance, risk, and the practical realities of building systems people can actually trust. This isn’t a debate about policy. It’s a look at how things really get decided, and what happens when they don’t.

1/3 - There is only one condition necessary for a business to exist: someone has to be willing to pay. This conversation steps away from the public narrative around UK AI and into the quieter reality of what it actually takes to build an early-stage AI company. Jack Perschke speaks plainly about funding, venture bias, domain expertise, and hiring in a global labour market, drawing a clear line between what looks coherent from the outside and what feels very different on the inside. There is no case study here and no success story to emulate, just a grounded account of how decisions get made when capital is selective, expectations are high, and execution matters. This is the first of three short films, and it sets up the wider question of who really defines the economics as AI moves from ambition into infrastructure.

For a long time, creativity, marketing, and software were separate, outsourced functions.AI is quietly collapsing those boundaries.In this conversation with independent brand consultant and creative Rob Paton, we explore what changes when execution becomes fast, cheap, and accessible to everyone.We talk about why traditional agency models are under pressure, why creativity still matters but shows up differently, and how AI-assisted tools are shifting work in-house. From branding and production to custom software and “vibe coding,” this episode looks at what happens when teams can build directly instead of buying capability.We also ground the discussion in a real case study from a heating installation business, showing how AI-assisted tools can remove bottlenecks, reduce cost, and change how decisions get made.This is not a debate about tools.It’s a conversation about judgement, responsibility, and what organisations choose to keep inside when making becomes easy.

This conversation is grounded in lived operational history. Simon Reid implemented the world’s first satellite operations automation system for EUMETSAT in 1990, and three decades later led the first commercial in-orbit AI automation platform with AWS and ESA, a system later used for real experiments by third parties. What follows isn’t speculation about AI in space, but insight shaped by decades of building systems that actually fly.Space is no longer a distant frontier. It is critical infrastructure.In this episode of BASELINE, Ian speaks with Simon Reid, CTO at Space DOTS, about why modern space systems have become too complex for humans to manage alone and what that means for satellites, data, trust, and decision-making.As Low Earth Orbit becomes more crowded and more contested, the risks are no longer just physical collisions. They include invisible threats such as radiation, signal interference, degradation, spoofing, and cascading failure across systems we depend on every day.Simon explains how the space industry is undergoing a quiet shift:from hardware-first thinking to data-first systems,from manual oversight to machine-driven inference,and from isolated sensors to aggregated, trust-based intelligence.

From 2026, the EU AI Act changes what accountability means for AI systems.For high-risk AI use cases, organisations will be expected to demonstrate traceability, record-keeping, and oversight. In practice, this means being able to show what an AI system did, why it did it, and who was responsible. Logs alone are no longer enough.In this episode of BASELINE, Martin Gibson explains why most organisations will not fail AI compliance because they lack data, but because they cannot reconstruct a defensible story across systems and time.We explore: • Why logs are not the same as evidence • What the EU AI Act actually expects in practice • Record-keeping, retention, and traceability for AI systems • Why AI compliance is a systems and infrastructure problem • How organisations get trapped by fragmentation and vendor lock-in • What it means to build AI systems you can defend under audit or investigationThis is a practical, operator-focused conversation for teams building, deploying, or governing AI in real organisations.

In this episode of BASELINE, filmmaker Archie Brooksbank reflects on storytelling, music, and the human connections that shape how we feel, remember, and act. Drawing on his work across film, sport, and culture, including projects involving figures such as Lionel Messi, Jamie Redknapp, and Lewis Hamilton, as well as collaborations with teams and brands including Red Bull Racing, Aston Martin, and McLaren, Archie explores why interaction, trust, and presence matter more than perfection. As AI and automation accelerate across the creative industries, this conversation asks a deeper question: what happens if we replace human experience with efficiency? This is not a discussion about tools or technology trends. It is a reflection on why storytelling has always been a shared human act, and why some things cannot be automated.

Artificial intelligence is already shaping organisations, governments, and public services. Many of the most important AI decisions are being made by default rather than by design.

In this episode of BASELINE, Ian speaks with Jonny Williams, Chief Digital Adviser to the UK Public Sector at Red Hat, about why AI feels confusing, why value feels elusive, and why leadership matters more than models.

This conversation explores the real architecture behind AI systems, how technology stacks align with national interests, and why transparency and openness are now prerequisites for trust. Jonny explains why AI is not primarily a technical challenge, but a leadership and operating model challenge, closer to the role of a COO than a CTO.

Topics include AI sovereignty, Wardley mapping, open versus closed models, national infrastructure, mechanised government, digital identity, and why organisations risk losing agency if they allow AI decisions to drift to technologists alone.