I’m reprising the post of 30 November 2013 called Voices without bodies. The blog continued the reflections in my book The Tuning of Place: : Sociable Spaces and Pervasive Digital Media published by MITPress in 2010. Here’s the original with some updates — assisted by ChatGPT. Question to Siri:“What’s The Wizard of Oz about?”Siri: “It’s about some Dorothy, her intelligent assistants, and her…More
The clone and I
I asked ChatGPT why I should use video cloning of my own appearance and voice. The response continues a thread in which I had already consulted about the reprise of my old 2013 posts. The response discloses some interesting points about time, identity, the labour of recording, and the reception of video content. My self-cloning…More
Thinking about walking
I’ve managed to link my cloned avatar (using HeyGen) to my cloned voice (via ElevenLabs) so that it reads the content of my blog from November 2013 called The benefits of walking (#170). The process was straightforward. The content pertaining to Aristotle’s four causes is fairly dense. For people interested, some material is easier to…More
My first video clone
My post #169 published on 9 November 2013 was titled Feeling free in flight and continued the theme of the cyborg. The cyborg is just one of a number of virtual hybrid human media entities. Another obvious example is avatars of players in video games. Recently, it’s been possible to combine video and audio presentations…More
Experts versus algorithms
I’ve been revisiting the post from 2 November 2013 about algorithms, drawing impetus from the book by Daniel Kahneman Thinking fast and slow. I titled my post Why experts are better than algorithms (post #168). I’ve just listened to the post recited slowly and with gravitas by the synthetic voice of John Rhys Davies (attached).…More
The “intellectual cyborg”
My post from 26 October 2013 called “What’s wrong with posthumanism” addressed the concept of the cyborg. By most accounts, the cyborg is an ambiguous category of human being whose functional performance is sustained by artificial supplementation and enhancements — from reading glasses to prosthetic limbs. The original proposition came from NASA scientists who identified…More
Metafiction and metaphor
Most fictional writing makes generous use of figurative language: hyperbole (exaggeration), anthropomorphism, simile. This much is obvious. As a reminder, here’s a fragment of imaginative writing from Italo Calvino’s Invisible Cities: “The slender stilts that rise from the ground at a great distance from one another and are lost above the clouds support the city.…More
The threat
“You told me you would start writing now you have time on your hands. You have plenty to write about since you came out of confinement.” I took that as a threat. If I didn’t give an account I could end up inside again. Then it dawned. My inquisitor was writing everything down. “Do you…More
Metafiction and melancholy
A metafiction typically reflects on the writing process of the author as the work is being written. Either fictional characters assume the role of the author of the piece, or actual authors weave themselves into the story and reflect on their own processes. Metafiction is an appropriate genre with which to explore melancholy. Scholars have…More
Pastimes revisited
I’m looking back at some of my blog posts from 2013. The posts canvas the topics: gaming culture, time management, design transitions, and online inference. As it happens, they each circle around the subject of evidence. Evidence is material for narrative, delay, aesthetic judgement, and inference. Let’s Play introduces the sociability of play, including video gaming…More
A tissue of citations
I’ve just completed an essay for a special edition of Architectural Theory Review in honour of my friend and former colleague Adrian Snodgrass, who passed away last year. Adrian introduced me to structuralism, semiotics and hermeneutics. In part of that essay I reference the writing of the literary theorist and cultural semiologist Roland Barthes (1915-1980).…More
How to explain art
Here is an audio of a sequence of blog posts I published in 2013. Some were written while on a trip back to Australia after an absence of 16 years. These reflections of a traveller later informed my books Mood and Mobility, Network Nature, and Derrida for Architects. My current challenge is to see what…More
Not everyone isn’t unhappy
I’ve just emerged from an interaction with ChatGPT reviving the following posts from 2013. I was interested in my early attempt to explain what we now think of as “confirmation bias.” The example I led with was of people arguing that nature settings can instil a positive mood. Scholars and everyday observers tend to select…More
What was I thinking!
Past writings, diaries, letters, course notes, publications and blog posts reveal what I was thinking 10 to 20 years ago and beyond—or do they? To read them now is to be reminded how much any authored text is steeped in the artifices of language, culture, and circumstance. There is no direct access to an original…More
Beyond attention
I’m still reviewing my blog posts from 2013. In keeping with the unstructured nature of blogging, I didn’t plan sequences of posts to follow particular themes. However, a thread does emerge from this sequence of six posts — that of attention. Soft fascination (138) introduces the theory that sustained concentration on a task induces “attention…More
Geometry and affect
I’m reprising some older blog posts from 2013 that consider vertigo, oblivion, melancholy, the motion of swings, and the emotional experience of urban spaces. Later on I drew this material together in my book Mood and Mobility (2016). With assistance from ChatGPT I’ve updated the sentiments of these posts under the rubric of geometry. I also attach…More
Am I in AI?
I’m a member of the Author’s Licensing and Collecting Society (ALCS). They collect money for “secondary uses” of publications – such as photocopies, digital reproduction and educational recordings. These rights bring in only small amounts of money, so authors don’t usually take the effort to collect them. Nor do they know how to do this.…More
Easy reading
I am reviewing my early blog posts Richard on the Holodeck and Shallow reading from February 2013. In the first post I noted that some people see literature (e.g. a novel by Charlotte Bronte) as a substitute for living the lives of the characters. In Hamlet on the Holodeck Janet Murray seems to suggest that…More
AI does history
I’m revisiting older blog posts. I’m up to the one titled “Loose ends,” which reflects on the nature of origin stories in the age of the Internet. The post from 2013, mentioned horsemeat detected in hamburgers, Derrida on the desire for beginnings, and Freud on red threads in naval ropes. I have recently traced a…More
Faroese chain dancing
I’m just back from a short break in the Faroe Islands. One evening I was standing in the reception area of our hotel when I heard the faint sound overhead of robust voices in unison accompanied by the slow rhythm of what seemed like feet stamping on the floor. The next day we were at…More
Vintage futures
I’m continuing the theme from the last post of looking back and looking forward. To that end I am reviewing the next set of posts authored in 2012. I start with the romance with digital technologies as they were then. When I wrote Vitruvius does steampunk in 2012, I was interested in how steampunk revelled in retro-futurism:…More
A digital time capsule
I’m looking back at old blog posts and publications. In 2012 I was also looking back to older studies, e.g. to 1994. See the 2012 post: “What’s a modem?” I’m indebted to ChatGPT for suggesting that the 2012 post served as a “time capsule.” In that post I revisited our 1994 study of computer-mediated communication…More
ChatGPT as trickster
As I continue to trawl through earlier blog posts I see that one of my 2012 posts followed a one week sojourn in Iceland. Iceland’s traditional narratives and mythologies draw on the activities of a pantheon of heroes, one of whom is the trickster god Loki. In that post, and subsequent publications, I referenced Lewis…More
Adaptive re-use of digital content
I’m in a cottage in the countryside. There’s an attic, a shed, a greenhouse, old gardening tools and furniture. Such legacy paraphernalia and economic necessity invite strategies of repair and re-use — practices that spill into the intellectual sphere. I’m re-using earlier blog posts. These posts were published between March and May 2012, beginning with…More
Pause for effect
I have been blogging about text-to-speech and voice cloning apps. I’ve also been turning compilations of my posts into audio files using Speechify suitable for podcasting. (See previous posts.) It seems the TTS (text-to-speech) tools I am using do not accommodate TTS HTML or other reliable means of introducing intonations and pauses into the readings.…More
2011 and all that
I am reviewing my early blog posts on technology, media and culture. The compilation here includes weekly posts between 9 April and 31 December 2011. The posts make reference to “actual events” that year. These include: Apple’s iPad 2 release; the impact of social media in the January “Arab Spring”; Osama Bin Laden’s death; the…More
2010 redux
This audio file (generated by Speechify’s synthetic voice) contains the first 41 regular weekly blog posts I published starting in 2010. I began the blog after the publication of my book, The Tuning of Place. I have just called on ChatGPT to set the context for the original content. My prompt: “This content was first published…More
Mood tags
Silent personal reading emerged in the 18th and 19th centuries, coinciding with the wider circulation of printed books and pamphlets, according to literacy scholars. Before this period, reading was typically performed aloud, even in solitude. Text-to-speech (TTS) software seems to be returning us to that read-out-loud practice. Our texts can be read to us by…More
Clonecasting
The term “clonecasting” often refers to copying an actor’s persona, presentation style, appearance, and voice to create visual and audio media content. Audiences might think they are seeing and hearing a particular actor in a film, but the actor’s presence is fabricated from digital models. Famous cases involve the reconstruction of the deceased actor Peter Cushing in Rogue…More
The enthusiastic clone
My first blog post appeared in 2010 and followed the publication of my book The Tuning of Place. I titled the post “Tuning as …” The Christmas eve that followed I produced a post “Silent Night.” Here is an excerpt from Silent Night in audio format. It runs for 2 minutes: That’s not me speaking,…More
The purloined voice
I’ve been busy pruning an overgrown laurel hedge. Apparently, the previous owner cultivated the entire leafy barrier from a single twig he had surreptitiously snipped from a hedge in a garden centre. His careful propagation from that twig—growing, dividing, and replanting—constitutes cloning. A quick AI-assisted search reveals that the verb “to clone” derives from the…More
Predicting AI “misconduct”
A recent headline in the Higher Education section of The Guardian said “Thousands of UK university students caught cheating using AI.” I could see that coming! It is as if some headlines (taglines, tweets and chyrons) are ready and waiting for a report, evidence base, study or authorised opinion to make them real. To help…More
Mnemonic infidelity
Training a large language model (LLM) starts with dividing a very large corpus of texts into basic units, i.e., recurring tokens (such as symbols and parts of words), and calculating the relative positions of tokens in the texts. These relationships are processed in a neural network to create very large numerical vectors that represent patterns…More
AI shopping
You know that a technology has gone mainstream when it starts to affect the consumer side of the retail trade. I first heard the term “AI shopping” on the BBC Radio consumer programme You and Yours: “Now, a growing body of evidence shows around a third of us are using AI to find products, plan…More
Local AI
Large language models (LLMs) are called “large” as they are trained on very large volumes of text data, about 570 Gb for ChatGPT 4.0’s base model (prior to fine tuning). It sounds like a lot, but my iPhone 16 Pro has a storage capacity of about 500 Gb. Other smartphone models have twice that. So…More
Mining Reddit
Reddit is a platform that supports themed community discussion forums called subreddits. I recently joined one of the over 100,000 active subreddits. It goes by the name “r/artificialintelligence.” Reddit is very popular, with over 95 million active daily users who post, reply and comment anonymously using aliases. Typically, a user of Reddit will post a…More
AI sycophancy
Everyone loves a compliment. Courtesy, politeness and good manners entered recently into the AI discourse via two threads. First is the apparent cost of non-essential words. In discussing the costs associated with the widespread adoption of conversational AI Sam Altman, the OpenAI CEO, remarked lightheartedly that “please” and “thank you” typed into a prompt increase…More
AI and the devil
There are good reasons to avoid the use of Ouija boards and other paraphernalia of supposed spirit communication. Apart from their associations with fraud and deception, some people see such devices and practices as “evil.” AI skeptics draw attention to the risks of bias, error, changes to work practices, putting people out of work, and…More
Esoteric AI
Tim Berners-Lee’s concept of the World-Wide Web took hold in the 1990s. Multi-channel text-based data communications transitioned to a singular access medium, the “browser,” for linked display, interaction and communication of text, graphics, sound and video content. As well as facilitating the dissemination of professional, educative, ordered and “rational” content, the world-wide web fostered an…More
Urban unconcealment
Here is an exchange with ChatGPT, published prematurely it seems, that demonstrates how the LLM seems to have adopted a profile of my interests from uploaded files and previous threads. Referring to my book “AI and Language in the Urban Context,” I asked: “In the book I make reference to ‘revelation.’ It’s from concepts by Heidegger…More
Beyond urban metrics
In AI and Language in the Urban Context, I make the case that cities are substantially linguistic entities, their social, cultural, and material dimensions shaped and sustained through conversations. As large language models (LLMs) exert increasing influence within public life, they not only automate services but contribute to these urban conversations. The AI Index Report 2025 that…More
Urban AI
“Everybody’s talking about AI!” This week’s joke is that the US Education Secretary spoke enthusiastically about the introduction of A1 into early schooling. A1 is a brown sauce, similar to HP sauce in the UK. Contrarily, the Stanford AI Index 2025 Annual Report shows that people are now more aware of AI, particularly large language…More
Mobile AI does fieldwork
My book came out this week: AI and Language in the Urban Context published open access (i.e. free to download) by Routledge. In that book I advance the case that places and spaces are enjoyed, appreciated, resisted, interpreted and even created through language — including conversations. LLM applications are evolving constantly. It’s feasible to think…More
The future of writing
A well-read academic colleague of mine once remarked that books that become important in the intellectual sphere are generally “well written.” That observation could apply to novels, fiction, non-fiction, popular science, academic books or any other genre. The judgement about what makes writing “good” presumably includes the quality of its arguments, messages, narrative content, but…More
Evidence and absurdity
In the previous posts I recalled five categories of design explanation. I used the innocuous example of deciding to insert a large window into a living room. I explored how a client/reader might regard various explanations as either sensible or absurd. I wanted to avoid the language of truth and falsity, arguing that decisions and…More
Absurdum explicandum
Following my previous past on different modes of explanation, I was interested in how an audience would decide that an explanation was inadequate, did not make sense or was full of errors. Avoiding the language of truth and falsity, I asked NotebookLM: “Please mirror these five categories of explanation with examples of explanations that most…More
Explain this …
In my previous post I explored how LLMs provide explanations of the “reasoning” by which they produce their outcomes (i.e. responses, answers, decisions). At least in their current iteration, platforms such as ChatGPT are tuned to generate an explanation independently of the process by which they came up with their response in the first place.…More
The meta-reasoning illusion
The ability of an LLM to provide an account of the processes by which it came up with its results, as in the inventory table I showed in the last posting, is indeed convincing. But apparently it is an “illusion.” I can’t yet find an article on the subject, so I have come to rely…More
AI explains itself (again)
I’m just back from a holiday in Malta. Two AI events entered my awareness. The first was meeting an American language teacher who said she had used AI to find her a “charming B&B” just outside Valletta. The second event was news that Elon Musk had sent an email to all US Federal employees to…More
Mobile AI and place
Some smartphones incorporate AI features. I recently acquired the latest iPhone that now supports “Apple Intelligence.” The operations are not yet seamless, but that development encouraged me to think about how conversational AI might inform my experience of a place. After all, location aware mobile phones track position, journeys, weather conditions, images and sounds of…More
Inferring a book from its index
I’m busy creating the index for my current Routledge book: AI and Language in the Urban Context: Conversational Artificial Intelligence in Cities. As yet I have had little success in deploying AI to create an index for me. ChatGPT will however create a book proposal from an index. It seems that ChatGPT called on assistance…More
Ubiquitous AI
“You know, we always wanted this vision of AI that is too cheap to meter. And having a really good free tier and all these things, it just makes AI sort of everywhere.” That’s a comment by AI engineer Sean (Swix) Wang on a recent Last Week in AI Podcast discussing how Google and other…More
AI reveals all
ME: “People seem to like me. They think I’m funny.” STUDENT: “I don’t know about that. But you look funny.” I’ll continue with the rare indulgence provided by conversational AI. It seems a writer can court AI, via NotebookLM, for personal analysis, as well as compliments, with a level of attention unmatched by scrutiny from…More
LLM plans a lesson
Google has developed a suite of AI learning tools based on their Gemini conversational LLM platform: “Grasp new topics and deepen your understanding with a conversational learning companion that adapts to your unique curiosity and learning goals.” These are explained at https://learning.google and some are available for trial. We might expect such systems to be…More
DIY professional service automation
Any of us may need non-routine specialist expertise on occasion, whether for legal contracts, disputes, claims, rental agreements, property purchases, employment conditions, pension arrangements, tax liabilities, inheritance, wills, probate, insurance options and claims, conditions of engagement, investment advice, warranties and guarantees, instruction manuals, planning guidance, building warrants, permissions, or statements about rights and regulations. A…More
Friction in the machine
I prompted Google’s NotebookLM with a request to draft a blog post on “romancing the artificial,” with content derived from my own writing uploaded to the platform, See previous post. It suggested that the sources offer a “nuanced perspective” on the topic, particularly as it relates to information technology. This phenomenon, characterized by a fascination…More
Coming to terms with AI
I uploaded a collection of my own books and a few articles to Google’s NotebookLM, which deploys RAG (Retrieval-Augmented Generation) techniques to interrogate any corpus. I entered the books in chronological order, as I was interested in how my ideas and themes might have changed over the period of their writing (3 decades!). I prompted…More
LLMs and the web
Early web browsers (e.g. Mosaic in the 1990s and its successors) relied on centralised servers that pre-processed (“crawled”) web pages to identify key terms, words and groups of word. The servers would then add these terms to an index optimised for quick lookup, linking the terms to the URLs of the pages in which these…More
The Quid Pro Coaster
Designers, writers, and illustrators in general are skilled at drawing inspiration from just about any source. As I have shown in previous posts, LLMs seem capable of something similar, especially with source texts that are colourful, include spatial cues, characters, situations and accounts of interesting experiences. But how do they handle dull, prosaic, non-spatial texts,…More
Dialectic between the verbal and the visual
In the 1980s, the architect, theorist and teacher Bernard Tschumi penned an essay “Spaces and Events,” which later appeared as a chapter in his influential book Architecture and Disjunction. In it he outlines an approach to architectural design, at least in a studio teaching context, that elides the literary with the pictorial, to the extent…More
From text to image via LLM
Text and writing are important components in creating architecture. To put it more strongly: text is deeply intertwined with the production of architecture, serving as more than a mere communication tool. Text impacts design thinking, theory, history, and the way the built environment is constructed — materially. As in my previous posts, I’m seeding this…More
AI briefs an architect
Inspired by the game design scenario of my last post (Share your expertise) I asked NotebookLM how the ideas outlined in the collection of sources (e.g. my own published writing) could inform the creation of a physical architectural work. Here’s what it said: Informing Architectural Design The sources provide a wealth of concepts and frameworks…More
Share your expertise
Google’s NotebookLM utilizes the RAG methodology for document interrogation, allowing users to gain insights from multiple texts simultaneously, while the University of Edinburgh has introduced its ELM built on ChatGPT for educational use. Both tools enhance understanding in specialized fields by enabling easy document uploads and tailored responses, benefiting users in architecture, law, and healthcare.More
Smashing the context window
In natural language processing LLMs, the “context window” is the range of text around a target word or phrase that constrains the sequence of text processed in calculating positional and attentional encoding, and hence prediction of the text that follows. A larger context window allows the AI model to capture more subtle relationships and dependencies…More
AI profiles you
In 2009, an academic received feedback through a 360-degree process, similar to analyzing their blog content with AI. The blog, started in 2010, includes 741 posts exploring technology, culture, and design. Using AI, the author inferred insights on personality and product interests, reflecting their intellectual pursuits and the evolution of digital analysis.More
Post-creative AI
There’s a phase of production after an author completes a manuscript (draft or final) when the author might need to compile an index, summaries for publicity, and other “extra-authorial tasks”? The main creative task of writing is complete and the tasks that follow could be undertaken by editors or others, or even an AI. In…More
Easy reading
I continue a discussion of my manuscript AI and Language in the Urban Context: Conversational AI in Cities (Routledge). See previous post: Too much feedback. I asked Google’s NotebookLM, “Would the average reader have any difficulty following the technical/mathematical explanations?” The AI responded: Readability of Technical/Mathematical Explanations The sources [actually, the manuscript] suggest that the…More
Too much feedback
Google has introduced an AI-driven document analysis platform called NotebookLM. You upload a document to the NotebookLM server via its website (notebooklm.google.com). The LLM generates a summary of the entire document and suggests some questions of the kind you see in textbooks that test a reader’s comprehension of the content. These are not generic questions,…More
AI fever dreams
Reinstating Panpsychism: Cities and the Machinery of Mind in the Age of AI City: A sprawling, feverish organ of matter and thought, a monstrous convergence where steel skeletons fuse with the blood and minds of the masses. It pulses with an energy both alien and intimate, a grand machinic beast, dreaming its electric dreams. To…More
Chain of thought
A report by OpenAI on its recent release of ChatGPT o1 asserts: “Similar to how a human may think for a long time before responding to a difficult question, o1 uses a chain of thought when attempting to solve a problem. Through reinforcement learning, o1 learns to hone its chain of thought and refine the…More
AI explains itself
I’m catching up with Yuval Harari’s latest book Nexus: A Brief History of Information Networks from the Stone Age to AI. He alerts readers to the challenge posed by the spread of opaque algorithms that are increasingly responsible for deciding on our behalf: bank credit, purchase choices, health, governance. In the face of this obfuscation…More
A jolly AI adventure
I asked ChatGPT to rewrite some of my glossary entries in the style of Enid Blyton. I said, “Please provide a rewrite of the following glossary entries, but in the upbeat style of Enid Blyton’s Adventure Series featuring Jack, Philip, Dinah and Lucy-Ann, together with Jack’s intelligent and talkative parrot Kiki.” Always ready for a…More
A Dubliner’s guide to AI
I asked ChatGPT to revise some of my AI glossary entries in the style of James Joyce’s Ulysses. It took a bit of prompting for this highly creative output, though the AI had trouble being ungrammatical. GPT (Generative Pre-trained Transformer): Original: GPT is a type of AI model designed to process and generate human-like text. It…More
AI in the apocalypse
Reza Negarestani wrote Cyclonopedia: Complicity with Anonymous Materials. See my post The twist of the pen. The ChatGPT training corpus seems to have picked up his interesting, apocalyptic writing style . Here are some “straight” AI glossary terms I am compiling, each followed by ChatGPT’s attempt to reconstruct them in the style of Cyclonopedia. Feed…More
AI through the looking glass
I invited ChatGPT again to configure elements of my glossary. See previous post A critical AI dictionary. This time I prompted the AI to reword some of the definitions in the manner of Lewis Carol’s Through the Looking Glass. Dall-e: A generative graphics product accessed via ChatGPT and other LLMs. DALL·E uses a transformer model…More
A critical AI dictionary
I am compiling a sensible glossary of AI terminology. I gave each to ChatGPT with the prompt: Please rewrite these glossary items in the style of Georges Bataille’s Critical Dictionary. I follow my own wording with ChatGPT’s response. Agent: An autonomous entity capable of taking data from its environment and performing actions to achieve specific…More
Curate your AI
Working against time constraints an author draws on productivity tools to speed the writing process. Back in the day, desktop word processing changed the game for most of us. I wrote my PhD and first book with MacWrite on a Mac Classic. Before that I probably used something like a Brother desktop computer, and before…More
Choose your critic
My last post looked at the potential of a large language model such as ChatGPT to operate as a critic. Developing further on that theme, it occurred to me that such a review may not always be in step with the author’s target readership. So, I presented ChatGPT with a 4,000 word book chapter titled…More
AI as critic
I listened recently to a radio program in which a collection of prominent entrepreneurs discussed some of the key examples this year of overhyped marketing. As an aside, one of the participants mentioned the disturbing error rate in ChatGPT’s responses to questions of fact, e.g. name ten distinguished alumni of the Fitzwilliam College, Cambridge? It’s…More
Meaning and attention
Understanding and misunderstandings in conversation often stem from emphasis. People tailor their responses based on where their conversational partners place emphasis. Attention distributions play a crucial role in inflecting responses in dialogue. Emphasis influences what comes next in a conversation, shaping the interaction between speaker and listener. Text-only conversational exchanges rely on context without additional cues. Attention, as demonstrated by LLMs, significantly affects the platform’s text generation capabilities and conversation continuation.More
Extending large language models
I’m in the process of identifying parallels between urban semiotics and large language models (LLMs), arguing that core facets of language competence parallel aspects of urban life, experiences and processes. We can identify eight core functions that contribute to the success of the Transformer model of LLMs, as deployed in ChatGPT and other chatbot platforms.…More
Urban scripts and contest
My previous post about scripts, language, and AI invites further reflection the role of scripts in the urban context. Theatrical-style scripts encompass the predictable and patterned ways in which individuals interact in urban settings. Think of how passengers might be expected to wait for those leaving the carriage before entering, give up their seat for…More
Scripts in the city
LLMs exhibit strong script-writing capabilities, crafting roles, settings, and dialogues based on extensive training. While they may not fully compose three-act plays, they excel in simulating dialogue and can assist in script creation. Scripts play a vital role in AI and cognitive science, offering efficient ways to convey information and understand various contexts, echoing the importance of scripts in language, cognition, urban contexts, and AI script writing.
“The words made me do it”
In Neo-Victoria, Detective Hayes confronts Dr. Marlowe, who insists that ancient alien language compelled her to commit murder. This fictional scenario reflects the power of language to influence actions. Language, as a generative force, shapes urban dynamics and societal behavior, as explored by influential thinkers like Foucault and Heidegger. Orwell’s 1984 exemplifies language’s capacity to manipulate behavior.More
Must AI speak?
AI is subject to critique from many quarters. See posts Generalised AI as existential threat, and Learn to talk to your AI. We don’t need to draw on philosopher Jacques Derrida to articulate the limits and challenges of AI, but here I want to identify a Derridean-style argument that impinges on my thinking about AI-based…More
AI and deconstruction
The early days of the internet saw scholars like George Landow and Jay Bolter draw parallels between hypertext and language. They utilized Jacques Derrida’s philosophy to explain the success of large language models, emphasizing how written text precedes speech. Derrida’s concepts of intertextuality, différance, authorship, and active reading resonate with the functioning of these models.More
Derrida on AI
The radical philosopher Jacques Derrida challenged the conventional belief that speech precedes writing, arguing that writing is fundamental to language’s structure and meaning. Current large language models, reliant on text, support Derrida’s theory, emphasizing the enduring and analyzable nature of written language compared to the ephemeral and context-dependent qualities of speech. (Word count: 50)More
Voice chat
The latest incarnation of large language models (e.g. ChatGPT) generates convincing spoken responses to spoken input. You can talk (as well as type) to your AI, as if on a smartphone. It’s called “voice chat,” and I used OpenAI’s new voice chat app on my smartphone to enter into a conversation clarifying certain aspects of…More
AI scripts a debate about urban glare
I’m testing the ability of large language models (LLMs), like ChatGPT, to conduct AI to AI conversations that in some way parallel how two people might work through a problem together. I’ve discovered that it’s not necessary to contrive a condition where two AIs talk to each other as in my previous post, or where…More
Manufacturing debate
I used copy-paste to implement a conversation between chatGPT and Claude. The two AI models were primed with the same warning about an apocalyptic future of AI-AI interactions. They agreed on everything and simply added further facts and opinions about warnings, the need for caution, legislative measures and more research on how to make AI…More
Pond algae: an inter-AI conversation
What happens when AI chatbots talk to one another? There are numerous transcripts, videos and reviews online of such inter-AI conversations — some real, fake or doctored, sometimes to demonstrate the shortcomings or absurdities of conversational AI. I am interested in whether such inter-AI conversations can lead to anything like collaborative problem solving. As a…More
Thought as conversation
Cogitating, rehearsing ideas in the mind, is a highly advantageous byproduct of our ability to converse with one another. That is, the ability of human beings to think things through silently and privately has developed along with the human ability to communicate with one another in language. Some theorists even assert that conversation comes first;…More
Can an AI only think “fast”?
In his popular book Thinking, Fast and Slow, Daniel Kahneman identifies how any individual is capable of making snap judgements, and with obvious advantage. That kind of “fast” thinking is necessary and appropriate in many cases. Jumping out of the way of an oncoming car, reaching for the chicken at a finger buffet, or raising…More
How to co-create with your AI
People often remember better, or differently, when in the company of others. A reading of sociologist Maurice Halbwachs (1877-1945) and science writer Israel Rosenfield (1939-) supports the collaborative aspects of remembering, recalling, interpreting and, in terms often used in logic and language studies — generalising. Conversation provides the primary demonstration of this capacity to recall…More
Neural networks that recall
It looks as though LLM technology can enhance web search without conceptually overhauling the web search methodology. (See previous post: AI versus web search.) But is there a way that neural network retrieval methodologies can be used as a substitute for explicit indexing? After all, the human capacity to recall does not rely on indexing.…More
AI versus web search
Discussions of AI in web search are often overtaken by concerns over the risks associated with automated user profiling (see post Surveillance capitalism and its discontents). Setting aside that raft of concerns, it’s worth reviewing claims about how AI facilitates efficient and helpful web search. AI enabled search Gemini (https://gemini.google.com/) is the name of Google’s…More
AI ethical filters
Claude.ai is a conversational platform using a large language model (LLM) developed by the Google-supported AI company Anthropic. It’s a rival to the popular ChatGPT. Claude’s self-claimed selling point is “I’m an AI assistant created by Anthropic to be helpful, harmless, and honest.” I tried Claude last year but lost interest when it told me…More
From CAD to AI
The complex relationship between cybernetics and semiotics outlined in previous posts helps me at least identify major differences in thinking about cities, and the lack of comprehension by adherents to one with the other. In part it’s a distinction between a calculative approach to cities, and a more humanistic orientation, and by “humanistic” I mean…More
Cybersemiotics
When it comes to the practical applications of AI, in particular urban AI, scholars look to historical legacies. As I reviewed in the last post, urban researchers such as Federico Cugurullo see the introduction of AI as mainly under the influence of cybernetics. That stems from the work of Norbert Weiner and his book Cybernetics,…More
Urban AI and cybernetics
If you want to investigate how two entities relate to one another then it can be helpful to introduce a third that is somehow relevant to both. Architecture and cookery can be conjoined via the concept of taste — that kind of thing. In my blog posts to date I present language as providing a…More
Semantic geocoding
In the previous post I examined the what3words proprietary geolocation/geocoding tool. With that system the 3 words are pre-assigned randomly from a lexicon of acceptable words. There are good reasons for detaching the 3 words from the content of the grid squares. If the 3 words referred to the content of the squares then map…More
Word-based geocoding
The tomb of Oscar Wilde is at at Père Lachaise Cemetery in Paris. The What3Words (W3W) address is ///crawler.falls.gossip. Tombs are good objects to locate via W3W geocoding as they typically fit within a 3×3 meter grid system. The cell immediately to the right of ///crawler.falls.gossip is geocoded as ///bottled.premises.beards. So the code does not…More
What’s wrong with pluralism
I’ve had many conversations with chatGPT on subjects as diverse as writing computer code to comparing and contrasting ideas from philosophy. The performance is impressive. It is a little like talking with an extremely well-informed and patient tutor. It is also unlike that kind of pedagogic human-human interaction in several respects. It becomes obvious to…More
Urban dissensus
To achieve consensus is a laudable aim of any decision-making body. Dissensus is the opposite: a condition of widespread dissent. Forays into the philosophy of Jacques Lacan (1901-1981) support the role of dissensus within communities, including urban communities. I examined the relevance of Lacan to digital media back in 1999 in my book Technoromanticism. I…More
City stories
I am investigating the potential role of conversational AI in the city. Some planning theory asserts an inevitable link between words and urban space. Planning theorist Lieven Ameel summarises this state of professional urban intervention in a 2017 article. The paradigm shift from a top-down kind of planning towards a more dialogic, participatory and discursive…More
AI and communicative action
At their best, developments in cities are informed by consultation with large numbers of residents and stakeholders. In his summary of the state of planning theory in 2005, Phil Allmendinger wrote: “Recent revisions to UK planning process predicated on inclusion and consultation perhaps add an addition to this mix, that of stakeholder values and desires…More
Urban inhibition
A group of students meets casually to relax and chat. One of the lecturers enters the circle. The conversation becomes stilted. Silence descends. The presence of any particular individual can excite conversation — or inhibit it. There are complicated dynamics in play here involving power, familiarity, social norms, and relationships within groups. That one agent…More
Conversation-centric computation
In the previous post I alluded to some of the challenges of encouraging large language models (LLMs, e.g. AI chatbots such as chatGPT) to communicate with one another in ways that are oriented to some task, and that don’t settle on merely exchanging platitudes. In spite of their instant access to parameters trained on vast…More
Societies of AI minds
In the 1980s, AI pioneer Marvin Minsky authored the book The Society of Mind, which outlined how we could think of the human mind as so many communicating agents: “Each mental agent by itself can only do some simple thing that needs no mind or thought at all. Yet when we join these agents in…More
AI agents
The sociologist Bruno Latour (1947-2022) and others who subscribe to actor network theory (ANT) are keen to admit objects and things as well as people into the field of social study. Hammers, kettles, baskets and remote controls [71] are participants, actors and agents in networks of relations. As well as providing settings for human actions,…More
An AI focus group
Before widespread digitalisation, I recall borrowing books from the library that were graffitied with penciled underscores and multi-coloured hi-lighter markings indicating what various past readers thought important. In terms of my previous posts about attention in NLP models, such markups constitute a compelling record of human-based “multi-headed attention.” I described multi-head attention in my previous…More
Multi-head attention
My fascination with electronic sound production began when my father brought home a Grundig TS 340 reel-to-reel stereo tape recorder. Like many other recorders it had separate heads for recording and playback, which meant that you could play a recording, add new sounds and feed the combination back through the recording head in real time.…More
Spatial affinities
In natural language models (NLMs), semantic embedding vectors capture the position of a token in multidimensional feature space. The space is derived from word proximities in a natural language corpus and is derived by the automated adjustments to weights within a neural network model. NLMs can deploy these vectors in a number of ways, including…More
Attending to the city
In the book Network Nature, I explored how people attune themselves to the natural world — or at least, how we attune to that part of our spatial experience that we are inclined to describe as “natural.” We also attune to artifice, such as urban environments. Cities present a spectrum of stimuli that shape our…More
Hiding one AI generated picture inside another
Steganography is the art of concealing one image inside another. I discussed the basics of the technique in a post: Hiding one surface inside another. It involves bit-shifting. In an 8-bit image each pixel is represented by an integer in RGB (red, green and blue). With 256 shades of RGB that provides over 16 million…More
Evaluating your AI
I’ve been implementing small scale trials of automated natural language processing routines that deploy the same methods as ChatGPT, i.e., implementations of the so-called Transformer architecture. That requires training on sequences of words in an original source document, equating each word to a semantic encoding (i.e., a long vector of 30 floating point numbers sourced…More
NLP sequences and cycles
In spite of its esoteric mathematical intricacies, automated natural language processing (NLP) as in conversational AI, draws on at least one primal everyday phenomenon. I’m referring to the concept of periodicity, i.e., cycles, periods, rhythms, repetitions, etc. Periodicity is a major principle through which we understand time, temporality, ordering, and sequencing and permeates so much…More
Robot probes city grid
Sentences are of course sequences of tokens (or words), and one of the major tasks of large language models (LLMs) is to predict the plausible continuation of sentences from some start condition, such as a prompt in a chatbot dialogue box, taking account of the previous flow of tokens within a given context window. Several…More
AI learns ABC
Neural networks (NNs) as deployed in automated language processing (NLP) are good at identifying and reconstructing patterns in data. So a NN trained on a corpus of texts can identify words that are commonly grouped according to their proximity within sentences, e.g., we wouldn’t be surprised to find words (tokens) such as “building,” “services,” “construction”…More
Chatting with multi-modal AI
ChatGPT now supports image upload, which means that I can chat about pictures, a step change in its application to architecture and design. Here’s my first such conversation, with chatGPT responses shown as quotes. The platform has information about my interests in urbanism and AI via the Custom Instructions settings. So responses reference my interests.…More
Non-NLP models
As an NLP neophyte I thought the mechanisms of automated natural language processing, such as ChatGPT, esoteric and arbitrary. At best, the general pre-training transformer (GPT) model is structured as a series of mathematical functions that apparently work in simulating human language, but bear little relation to the processes by which language is actually spoken,…More
By the same token
Automated natural language processing (NLP) of the kind deployed in conversational AI typically breaks blocks of text not just into words, but into tokens. Tokens are strings of characters extracted from a corpus of texts — the set of texts an NLP system is trained on. Tokens are more efficient than storing just whole words.…More
The neural neighbourhood
In a previous post (AI makes AI) I described an an auto-associative neural network with an example using made-up data. Here’s a more sophisticated version with actual urban data. Consider this fragment of a city map. I overlaid a 100 metre grid carving the neighbourhood into 72 cells. Within each of these cells lies a…More
My first AI language model
Here is output from my first home grown generative language model. I trained it on my own writing (19,000 words), it extracted my vocabulary and it was able to produce nonsensical but better than random text, considering the simplifications and assumptions in the language model I used. “Unveils ancient daydream charles regulatory demands interactive aggregate…More
AI makes AI
One of the benefits of (and concerns about) large language models (LLMs) such as ChatGPT and its successors is that it can write computer programs, even AI programs. The concern is that it might start to do so autonomously without human intervention, but that’s for another post. GPT writes code I took advice from ChatGPT…More
Find yourself in AI
Guest post In the vast expanse of the digital universe, AI, particularly large language models (LLMs) like ChatGPT, have found a home, influencing our lives in countless ways. A lingering question, however, is whether we are inadvertently influencing these models back. In other words, could our writings have been part of the training data for…More
What’s next
Industry, education and everyday users are making increasing use of large language models (LLMs) driven in part by the prominence of ChatGPT and other AI tools. The technology is developing at a pace. The analysis of commentators, critics and legislators also gain traction as they evaluate the implications of the technology and seek to influence…More
Meaningful, imaginative, fluid
The recent escalation of interest in text-based AI coincides with our students completing their project work, dissertations, research degrees, etc. So it’s been a useful period to test the capabilities of platforms such as ChatGPT4 as a tutor — or perhaps as co-tutor. As yet there are no constraints in place that seriously impede its…More
Training your AI
The training data for a GPT model such as ChatGPT4 consists of many (hundreds of billions!) tokens harvested in sequence from various online sources, such as Wikipedia amongst many others. This source data is processed as a continuous stream independent of document boundaries. Tokenize The initial training task is to tokenize the entire corpus, identifying…More
More on automated recollection
Conversational AI platforms such as ChatGPT generate predictions for what should come next after a user inputs a question, statement, paragraph, or other text prompt. In early text prediction software, a simplistic language model might calculate the most likely word to follow a given input like “door,” based on pre-calculated statistical analysis of word co-occurrences…More
The human in AI
As critics contemplate the artificial in artificial intelligence, it’s worth considering the extent to which human agency plays a role in this technology, particularly in LLMs (large language models) and conversational AI. These technologies are invented, developed and improved by human beings, and they are fuelled by gigabytes of human generated texts. That much is…More
Why a neural network forgets
Conversational AI, such as ChatGPT, has limited capacity to recall the content of earlier conversations. OpenAI does not disclose all the details of its operations, but scholars estimate that ChatGPT4 can process and recall up to 10,000 words in a single session or thread. That’s a substantial improvement on earlier models, but it doesn’t ensure…More
AI textsurfing
Before digital text markup I would highlight key words in a difficult document with a coloured highlighter pen. Students with a more methodological inclination developed this into an art, with colour coding and supplementary markings and marginal comments. (I see that Staedtler imply that their pen users are textsurfers.) Now my markup practice is fairly…More
Can I use AI in academic writing?
I’ve been coaching masters dissertation students as they complete their final projects. I’m interested in large language models (LLMs) and their applications. At the moment, it’s easy to include the ChatGPT platform as a participant in one-on-one discussions with students. The subject matter of their projects is digital media, so familiarity with the applications, strengths,…More
Attending to more than one thing at a time
In an earlier post (Attention Scores) I considered how automated natural language processing (NLP) models attempt to simulate the way a listener or reader will focus on key words and groups of words in a sentence to decide how to respond to the sentence. I won’t repeat the calculation here. But recall that the automated…More
Confidential documents and conversational AI
Confidentiality is key in any profession, especially as it related to client-consultant relationships. I’m hard pressed to find confidentiality foregrounded in architectural codes of practice, but it is crucial in law and financial services. The Handbook of the Financial Conduct Authority, for example, states that a financial advisor (a “skilled person”) “may not pass on…More
AI and the Cryptographic City
My book Cryptographic City: Decoding the Smart Metropolis (MITPress) came out in May 2023. Here’s a summary taken from the MIT Press website. Cryptography is not new to the city; in fact, it is essential to its functioning. For as long as cities have existed, communications have circulated, often in full sight, but with their…More
I took my AI on a holiday
Here’s what we talked about, unedited and unabridged. It’s very upbeat! Illustrations are from my holiday in Mauritius June 2023. Me: Like many idyllic islands in the tropics, Mauritius has a dark past. ChatGPT: Indeed, like many places around the world, Mauritius has a complex history that includes both positive and negative aspects. While the…More
Welcome to the Apocalypse
Not everyone is averse to the prospect of a global AI-induced apocalypse. Catastrophizing circumstances and events caries a certain appeal to some, in particular those who identify with the status of a powerless underclass. Let social, political and economic systems fall! Let AI take over! I think here of those who identify as dispossessed, who…More
Generalised AI as existential threat
Many specialised AI applications are acceptably proficient in identifying people, animals and other objects in pictures, in searching databases, winning at chess and in many other areas invisible to most users, such as controlling factory production lines, navigating aerial vehicles, surveillance, medical imaging, diagnosis, and aspects of smart city infrastructures. That’s “narrow” AI. But general artificial…More
Attention scores
One of the important techniques in the new wave of highly successful generative natural language (NLP) models is the use of attention scores in neural network (NN) training. Here I continue the investigation I started in previous posts into how NLP works. To recap: a word in NLP models is represented as a point in…More
AI Armagedon
Should AI research be shut down? A group called the Future of Life Institute warns against the recent developments in generative AI platforms, prompted especially by advances in natural language processing, e.g. chatGPT. The open letter follows their similar warning of 2015. Both letters have several high-profile signatories who are involved in AI development. The…More
Words in order
Neural network researchers invented several methods that store and make inferences about the order of words in a sentence. The main method I will present here provides one of the components that undergirds the recent impressive performance of natural language processing (NLP) models known as transformer models. The method also resonates with my prior investigations…More
AI eschatophobia
It seems that everyone is talking about (or with) ChatGPT. The platform’s convincing conversational acuity and ability to synthesise disparate conceptual threads provides a vivid demonstration of AI’s potential. I’ve now read several accounts online where scholars, programmers, writers, musicians and artists use ChatGPT or similar as a creative companion to explore ideas — comparable…More
Fine tune your AI
My first blog posting appeared 25 August 2010 titled “Tuning as …” It followed the publication of my book The Tuning of Place. In the short posting I proposed: “Tuning might well be the metaphor of our age. ” I was thinking of smartphones and their role in moderating our relationships with environments. The term “tuning,”…More
Architecture in multidimensional feature space
In a previous post (Predicting proximity), I reviewed the NLP (natural language processing) operation of calculating the relationships between words in a corpus of texts. So the word “architectural” is closer to the word “urban” than “architectural” is to “culinary.” Very close word proximity could mean that one word can be substituted for another in…More
The invention of language
Automated natural language processing (NLP) technology is impressive, though I need to remind myself of its limits. It is easy to elide thoughts about its linguistic capabilities with a sense that it is on the way to mastery of language, and therefore human intelligence. In what follows I will follow the line that NLP is…More
Imperfect patterns
Training a neural network (NN) involves automatically adjusting numerical weights and thresholds (biases) to account for all input and output pairs presented to the NN. After training on these input-output patterns the NN should reproduce the appropriate output pattern when presented with any one of the input patterns. Note that it is not patterns that…More
Just one neuron
I revisited our earlier (1990) article on neural networks “Spatial applications of neural networks in computer-aided design.” Neural networks were novel in architecture and CAD. What follows is an update of the part of that article in starting to explain how neural networks function. Neural network layers Neural network (NN) models store information as numbers…More
Mandala and metaphor
I’m on sabbatical as I write this, and taking the opportunity to research AI in the urban context as I travel. Here I am in Central Java, at Borobudur for the first time. This trip was prompted by my early work with Adrian Snodgrass, with whom I joint authored the book Interpretation in Architecture in…More
Automatic pattern completion
In 1990 Arthur Postmus and I published an article about “spatial applications” of artificial neural networks (ANNs). In a more recent article, Gabriele Mirra and Alberto Pugnale at the University of Melbourne developed up-to-date applications of AI in design. They generously cited our article (amongst articles by others). I concur with their assessment of the…More
Who’s listening?
In every mind there is a higher function watching or listening to the inflow of sense data and monitoring its own thought processes. In his seminal book on the philosophical challenge of consciousness, Daniel Dennett argues convincingly against this proposition. He directs his criticism against those who presuppose that somewhere, conveniently hidden in the obscure…More
The next word
“We are so in sync we finish each other’s … sandwiches.” That’s a variation on a joke from several sources: The Simpsons, Arrested Development, The Good Place, etc. In some social situations where I can’t think of anything appropriate to say, I just start a sentence anyway, not knowing what comes next or how it…More
Predicting proximity
Tokens NLP (natural language processing) systems don’t necessarily consider only words as the basic components of prediction. They may break words into smaller units roughly corresponding to syllables, punctuation marks and even single characters. The general term for these textual units is “tokens.” An NLP system will include in its algorithmic workflow the construction of…More
Nothing beyond text
In explaining the political philosopher Jean-Jacques Rousseau’s (1712-1778) autobiographical book Confessions, the continental philosopher Jacques Derrida observed that there is nothing to support Rousseau’s claims about his relationships with his family members, other than what the text discloses. Derrida generalises this observation to the controversial proposition that any attempts to ground spoken or written assertions…More
Attention is everything
Attention is a key element in cognition. At our most thoughtful we direct attention to features in our environment that are most important to us at that moment. Attention can wander, of course, we daydream, and we can pay attention to inexistent things, memories and objects of the imagination. A lecturer will come to the…More
Learning to transduce
Human beings at their most rational are able to generalise from examples. If you stand under the shower head before turning on the tap then it is likely you will be dowsed with cold water before it gets to a temperature you are happy with. That if-then rule is a generalisation borne of a few…More
My tutor is an AI
In constructing this post I’ve been cross checking some of my explanations about neural networks with OpenAI’s chatGPT-3, which is a highly responsive chatbot available at the OpenAI.com website (free at present). I’m most familiar with spreadsheets. So that’s where I’ll start. You can think of a neural network as a matrix or grid (a…More
Intentional systems
In his book Consciousness Explained, the philosopher and cognitive scientist Daniel Dennett says that a study of hallucination, “will lead us to the beginnings of a theory — an empirical, scientifically respectable theory — of human consciousness” [4]. I’ve explored what some philosophers say about hallucination in previous posts and tried to relate that to…More
Chatting with an AI about urban inexistence
“Intentional inexistence” is a philosophical term adopted by the nineteenth century philosopher Franz Brentano (1838-1917) to indicate the commonplace human capacity to imagine things. We may suppose that inexistence refers to things that don’t exist (nonexistent things) or that are under demolition, but that presumes too much — or too little. As explained by Linda…More
Cryptographic City
The description of my next book is out with the cover design on the MIT Press website. The graphic designer says that the cover is in code, though I have yet to decipher it. Cryptography’s essential role in the functioning of the city, viewed against the backdrop of modern digital life. Cryptography is not new…More
Cryptographic index
To introduce cryptography as a theme in my Media and Culture class I provided each student with a customised string of code containing the name of an AI-themed film. The key to decode the secret message was the student’s unique 7 digit university identification number. If the encrypted string is uaroulojz and the student key…More