Dr Martín Raskovsky

She needed words for twelve pieces of art and could not find them. He asked one question at dawn and then walked on. This is a story about finding, set on the cliffs and beaches of Cornwall, where the tide turns whether you are ready or not.

Keep reading

She reached for an image. He reached for a symbol. Between them they named something they had been circling since the dawn on the dunes.

Keep reading

The question arrived, as M's best questions do, sideways and without warning.

Keep reading

W brought the question with him like something he had found in his pocket. He had not planned to. It arrived in the conversation without announcement, already there.

Keep reading

M was not stuck in the way she had been stuck before. This was a different kind of stuck. She knew the tools now. The tools were not the problem.

Keep reading

The dream had faded before she could reach for her phone. M was left with its temperature, its shape, and a carpet of Cornwall photographs spread across the floor near the fire. What she was looking for was not in any of them. Or not quite. Or not yet.

Keep reading

The stories in the M W Σ series - one original and five companions - were written for the reader who finds technical language about AI impenetrable. This essay is for the reader who finished the stories and wants to know what was actually happening underneath - the abstractions named, the technical terms placed against the story moments that carried them. Embeddings, hallucination, the context window, multimodality: the machinery behind M, W and Σ.

Keep reading

How does a machine reason? In November 2025, three days of building a semantic search engine burned away the mystery of how machines find similar things - revealing geometry underneath. This essay takes the next step: from finding to thinking, from geometry to reasoning - and asks how far that journey has actually gone.

Keep reading

The most powerful instrument for building a decentralised web is also the most centralised technology on the planet. This essay argues that this paradox is not a permanent condition - computing has broken this kind of stranglehold before, and history suggests it will again. The question is not whether AI will be decentralised, but what kind of rupture will make it possible.

Keep reading

An open letter to the Xibit artist community

Keep reading

What is the relation between two life passions - software engineering and photographic art - that appear to come from different worlds? After five decades of problem solving in code and years of transforming images into emotional expression, I have come to see them as the same activity wearing different clothes. Both involve traversing a dark tunnel toward an unknown destination, guided by intuition and experience, until a moment of illumination reveals what was hidden. The eureka moment - whether finding a misplaced dot in a system or discovering the emotional truth beneath the pixels - brings the same tremendous pleasure. Society likes its categories: engineer or artist, analytical or creative. My life is evidence that this is a false dichotomy.

Keep reading

Throughout history, technological advances have triggered fears of diminished human agency-fears that often prove justified. Yet each wave of technology contains not only the danger but its remedy: the same tools that enable control can enable liberation. This essay explores the concept of the reversal-the threshold at which technologies of exploitation become technologies of sovereignty. Using personal data as today's battleground, it argues that the difference between technological subjugation and empowerment lies not in the tools themselves but in education: understanding deep enough to see the reversal when it arrives, and to act on it.

Keep reading

A routine debugging session with an AI assistant revealed something unexpected: the bug we were hunting - a simple "greater than" that should have been "less than" - exposed a fundamental weakness in how AI generates code. The AI reasoned correctly in plain English but produced inverted logic in the actual program. This isn't a new problem. Forty-three years ago, my PhD thesis addressed exactly this gap between what we say we want and what code actually does. The solution I proposed then - formal transformation rules rather than verbal reasoning - turns out to be surprisingly relevant to understanding why AI coding assistants confidently produce plausible but wrong code.

Keep reading

After five decades in software development, I've learned that the most attractive technical journeys aren't about conquering complexity-they're about transforming the intimidating unknown into something elegantly simple. This week, I experienced exactly that: three days of building an AI-powered semantic search engine for artwork, where each step burned away another layer of mystery until what remained was not magic, but mathematics we could see, understand, and demonstrate.

Keep reading

This paper documents the emergence of a fundamentally new software development methodology observed during the design and implementation of a complex blockchain indexer system. Traditional software engineering has long mandated the separation of architectural design from implementation, driven by the economic reality that thinking was cheap while coding and changing code were expensive. This separation, while historically rational, created an artificial barrier that delayed architectural validation until implementation phases, often resulting in designs that failed when confronted with real-world constraints.

Through a detailed case study of solving database coordination issues in a blockchain NFT indexing system, we demonstrate how AI-enabled rapid implementation fundamentally alters these economic assumptions. When implementation becomes nearly instantaneous and architectural alternatives can be explored in parallel, the traditional design-first approach not only becomes unnecessary but potentially counterproductive.

We introduce "Architecture Through Implementation" (ATI), a methodology where architectural thinking occurs through concrete implementation rather than abstract specification. In ATI, multiple architectural solutions are implemented and validated in real-time, allowing empirical comparison of approaches that would traditionally exist only as theoretical designs. This process reveals edge cases, performance characteristics, and integration challenges immediately rather than after costly development cycles.

The paper examines how this approach integrates rather than abandons rigorous architectural thinking, making formal methods and systematic analysis more practical by applying them to actual implementations rather than abstract specifications. We analyze the skills and mindset shifts required for effective ATI practice, including rapid pattern recognition, implementation literacy, and collaborative prompting techniques.

Our findings suggest that AI-enabled development tools are not merely accelerating traditional development processes but enabling a qualitatively different approach to software architecture-one that collapses the artificial separation between design and implementation, resulting in more robust, empirically validated architectural solutions developed at unprecedented speed. This methodology may represent a fundamental shift in how complex software systems are conceived, designed, and validated.

Keep reading

The conversation around AI in software development has become frustratingly predictable. On one side, evangelists proclaim that AI will replace programmers, automate everything, and solve all our development woes. On the other, skeptics warn about job displacement, code quality degradation, and over-dependence on tools we don't understand. Both perspectives miss the nuanced reality that emerges from actual, sustained collaboration with these systems.

After five decades in software engineering-and recent intensive work with Claude AI on real-world projects-I've come to believe that the most interesting insights about human-AI collaboration exist in the messy middle ground that both camps tend to ignore.

Keep reading

The integration of artificial intelligence (AI) into software development has revolutionized how developers approach coding, debugging, and collaborative problem-solving. Through my own experience using ChatGPT as an AI-powered assistant, I have gained valuable insights into its strengths and limitations in the context of software development. This essay explores these aspects, offering reflections on how AI tools enhance productivity, where they fall short, and what this means for the future of human-machine collaboration in programming.

Keep reading

This essay redefines NFTs as digital assets, emphasizing their role as versatile licenses on the blockchain rather than mere digital collectibles. It explores how digital assets empower artists and creators to monetize their work in innovative ways, enabling applications in both digital and physical realms. Platforms like Xibit on the Vitruveo blockchain simplify artist onboarding while offering flexible licensing options for art, merchandise, and more. The essay also highlights how Verse, another platform on Vitruveo, is transforming the film industry by providing innovative tools for funding, licensing, and rights management through digital assets. By focusing on their practical utility and creative potential, this essay positions digital assets as pivotal to a new cultural and creative renaissance, bridging the digital and physical worlds and reshaping how we value and engage with art and media.

Keep reading

Photography has long stood at the crossroads of technical skill and artistic expression. Its legitimacy as an art form has been debated, with many acknowledging its artistic merits while others see it as merely a technical exercise. This essay defends photography as an art form, exploring its core artistic elements and addressing common misconceptions. It then examines the impact of generative collage tools, which expand the creative possibilities for photographers and further solidify their role as artists. The essay concludes by revealing an unexpected twist that challenges contemporary prejudices about the integration of advanced tools in the creative process.

Keep reading

The digital era has revolutionized the accessibility and distribution of art, bringing both opportunities and challenges. To protect digital artworks from theft and unauthorized use, various measures like watermarking and low-resolution images have been employed. However, these measures often compromise the observer's experience. The rise of AI and advanced technologies such as blockchain, Web3, and KYC protocols presents new solutions that promise to balance protection and visual integrity. This essay explores the impact of traditional protective measures, the transformative potential of new technologies, and the emergence of blockchains like Vitruveo which integrates these advancements to enhance the digital art ecosystem.

Keep reading

The distinction between skill or craft and artistic ability is a nuanced topic that pervades various forms of artistic expression, from music to the visual arts, and even to emerging technologies like AI.

Keep reading

Throughout history, technological advancements have often sparked fears among traditional practitioners that their crafts would be rendered obsolete. However, these concerns have largely proven to be unfounded, as each new innovation has ultimately enriched rather than replaced existing practices.

Keep reading

Artificial intelligence (AI) has revolutionized the world of artistic expression, offering new tools, techniques, and possibilities for creators to explore. However, the integration of AI into the artistic process raises profound questions about authorship, authenticity, and the role of technology in creativity. While AI-assisted artistic expression presents exciting opportunities for pushing the boundaries of artistic innovation, it also challenges traditional notions of artistic practice in the digital age.

Keep reading

In the realm of contemporary art, the integration of artificial intelligence (AI) with photography has sparked intriguing debates and controversies reminiscent of historical tensions surrounding the emergence of photography as an artistic medium. This essay will explore the parallels between the skepticism faced by both photography and generative art, delve into the complexities of copyright issues, and examine the creative possibilities when artists use their own photographs as references in generative processes.

Keep reading

The introduction of smart contracts in blockchain technology can be seen analogous to the groundbreaking concept of stored-program computers in the mid-20th century. Both concepts empower systems with programmability and flexibility, ushering in new possibilities and shaping the future of technology in their respective eras.

Keep reading

In the intricate dance between art and technology, photography stands as a testament to the ever-unfolding narrative of creative expression. Dr. Martín Raskovsky, a talented photographer based in the UK, seamlessly blends tradition and innovation in his artistic journey.

Every image in Martín's collections is a result of his photography, transformed through a dynamic interplay of digital manipulation, photographic collage, and generative processes.

Keep reading

It is the year 2023, the gold rush during the initial NFT frenzy in the early 2020s has long gone. Are NFTs dead ?

Keep reading

In the vast expanse of the internet, commonly referred to as the web, a curious labeling trend has emerged: Web1, Web2, and Web3. But what exactly do these designations signify?

Keep reading

I have just started in the NFT world - only weeks ago. This is my take on NFTs so far.

Keep reading