What do you think?
Rate this book
Twitterbots: Making Machines that Make Meaning
The world of Twitterbots, from botdom's greatest hits to bot construction to the place of the bot in the social media universe.
Twitter offers a unique medium for creativity and curiosity for humans and machines. The tweets of Twitterbots, autonomous software systems that send messages of their own composition into the Twittersphere, mingle with the tweets of human creators; the next person to follow you on Twitter or to “like” your tweets may not a person at all. The next generator of content that you follow on Twitter may also be a bot. This book examines the world of Twitterbots, from botdom's greatest hits to the hows and whys of bot-building to the place of bots in the social media landscape.
In Twitterbots, Tony Veale and Mike Cook examine not only the technical challenges of bending the affordances of Twitter to the implementation of your own Twitterbots but also the greater knowledge-engineering challenge of building bots that can craft witty, provocative, and concise outputs of their own. Veale and Cook offer a guided tour of some of Twitter's most notable bots, from the deadpan @big_ben_clock, which tweets a series of BONGs every hour to mark the time, to the delightful @pentametron, which finds and pairs tweets that can be read in iambic pentameter, to the disaster of Microsoft's @TayAndYou (which “learned” conspiracy theories, racism, and extreme politics from other tweets). They explain how to navigate Twitter's software interfaces to program your own Twitterbots in Java, keeping the technical details to a minimum and focusing on the creative implications of bots and their generative worlds. Every Twitterbot, they argue, is a thought experiment given digital form; each embodies a hypothesis about the nature of meaning making and creativity that encourages its followers to become willing test subjects and eager consumers of automated creation.
Twitter offers a unique medium for creativity and curiosity for humans and machines. The tweets of Twitterbots, autonomous software systems that send messages of their own composition into the Twittersphere, mingle with the tweets of human creators; the next person to follow you on Twitter or to “like” your tweets may not a person at all. The next generator of content that you follow on Twitter may also be a bot. This book examines the world of Twitterbots, from botdom's greatest hits to the hows and whys of bot-building to the place of bots in the social media landscape.
In Twitterbots, Tony Veale and Mike Cook examine not only the technical challenges of bending the affordances of Twitter to the implementation of your own Twitterbots but also the greater knowledge-engineering challenge of building bots that can craft witty, provocative, and concise outputs of their own. Veale and Cook offer a guided tour of some of Twitter's most notable bots, from the deadpan @big_ben_clock, which tweets a series of BONGs every hour to mark the time, to the delightful @pentametron, which finds and pairs tweets that can be read in iambic pentameter, to the disaster of Microsoft's @TayAndYou (which “learned” conspiracy theories, racism, and extreme politics from other tweets). They explain how to navigate Twitter's software interfaces to program your own Twitterbots in Java, keeping the technical details to a minimum and focusing on the creative implications of bots and their generative worlds. Every Twitterbot, they argue, is a thought experiment given digital form; each embodies a hypothesis about the nature of meaning making and creativity that encourages its followers to become willing test subjects and eager consumers of automated creation.
360 pages, Hardcover
Published September 11, 2018
49 people want to read
About the author
Tony Veale
10 books2 followersTony Veale is Associate Professor of Computer Science at University College Dublin.
Ratings & Reviews
Friends & Following
Create a free account to discover what your friends think of this book!
Community Reviews
Displaying 1 - 7 of 7 reviews
Read
September 19, 2018Considering the self-evident fact that Twitterbots are very likely going to bring about the end of human civilization (and neither want to do that nor know they're doing it nor even notice once they've done it), it's extremely unlikely that a book about these little self-replicating engines of deception and manipulation could be ENJOYABLE, and yet this book manages to be! My full review is here: https://openlettersreview.com/open-le...
September 9, 2018
This is an odd one. It's an in-depth look at Twitterbots - applications designed to post on Twitter making human-like pronouncements. We start with a really interesting, if highly verbose introductory section about these programs, introducing me to lots of examples I hadn't come across. I particularly liked the historical examples of condensing a message with humour (and a bit of intellectual oddity) in the Latin punning telegraph messages that seem to have been briefly popular amongst the British establishment in the mid-nineteenth century. For example, in 1856, when the British annexed the Indian province of Oudh (rhymes with loud), the governor-general sent the message 'Vovi.' This means 'I have vowed'... which sounds distinctly like 'I have Oudh.'
Then we plunge into the mechanics of Twitterbot construction. Tony Veale and Mike Cook, two British/Irish academics writing with a distinctly transatlantic style, give us detailed guidance on simplistic bot production using simple templates and collections of words or phrases, and also explore some more sophisticated methods using Java.
The book then turns into an exploration of the way that a program can build language structures - the kind of categorisation needed to provide a tweet builder with a 'Lego set' of appropriate words and phrases and the way it's possible to use some of the classic (if slightly hackneyed) academic breakdowns of stories, plots and characters to generate a pseudo-story format out of components (if you've read one of those 'how to write a novel' books that breaks down a hero's journey or whatever, into standard stages, you'll know the kind of thing).
What we end up with is a mix of an investigation of the world of Twitterbots, a how-to manual to get started writing your own code for them, and an analysis of just how much story can be broken down and built up in this component fashion. I found the presentation over-wordy and rather too academic in approach - but nonetheless interesting.
What the authors don't do - perhaps because they themselves are part of this scene - is some kind of psychological analysis of the nature of adults spending a lot of time on what is, in the end, the generation of rather weak snippets of writing that are only occasionally mildly amusing. The only reason the resultant tweets are of interest is because they're machine produced. It's rather like Samuel Johnson's dog walking on its hind legs: it's not done well, but you are surprised to find it done at all.
Then we plunge into the mechanics of Twitterbot construction. Tony Veale and Mike Cook, two British/Irish academics writing with a distinctly transatlantic style, give us detailed guidance on simplistic bot production using simple templates and collections of words or phrases, and also explore some more sophisticated methods using Java.
The book then turns into an exploration of the way that a program can build language structures - the kind of categorisation needed to provide a tweet builder with a 'Lego set' of appropriate words and phrases and the way it's possible to use some of the classic (if slightly hackneyed) academic breakdowns of stories, plots and characters to generate a pseudo-story format out of components (if you've read one of those 'how to write a novel' books that breaks down a hero's journey or whatever, into standard stages, you'll know the kind of thing).
What we end up with is a mix of an investigation of the world of Twitterbots, a how-to manual to get started writing your own code for them, and an analysis of just how much story can be broken down and built up in this component fashion. I found the presentation over-wordy and rather too academic in approach - but nonetheless interesting.
What the authors don't do - perhaps because they themselves are part of this scene - is some kind of psychological analysis of the nature of adults spending a lot of time on what is, in the end, the generation of rather weak snippets of writing that are only occasionally mildly amusing. The only reason the resultant tweets are of interest is because they're machine produced. It's rather like Samuel Johnson's dog walking on its hind legs: it's not done well, but you are surprised to find it done at all.
December 30, 2020
Twitterbots: Making Machines That Make Meaning provides a thorough introduction to Twitterbots, the main categories they fall into, and how they’re made. The authors include lots of links to resources for anyone who wants to build a Twitterbot or investigate the impressive resources the community has made. The most interesting sections of the book were the first several chapters. Unfortunately, I found the later chapters tended to drag a bit and go on about topics longer than was necessary.
My primary frustration with this book is that it sometimes ignores the environment of Twitter and instead comes close to treating it as nothing more than a platform the bots post on. This is strange to me, as the mash-up and juxtaposition function of Twitterbots is perfectly at home on Twitter itself and the internet more generally. The authors sometimes discuss this, particularly when it comes to how some users will use an opportunity to find a way to use hate speech or crude jokes when they interact with a bot or brand, and more generally in the last chapter. However, it sometimes felt like the authors forgot about it, particularly when discussing how Twitter bots can make the familiar feel unfamiliar via unexpected mash-ups or words or content, when this is in fact a huge part of the Twitter experience already.
Perhaps the authors’ take on Twitterbots is the result of the particular perspective of the Twitterbot community and those around it. The authors quote Erin McKean: “[Twitterbots] take you outside your news-outrage-and-sandwiches timeline to give you a minute to look at the world in a new way.” I’m sure a lot of people would not describe their timeline this way, and instead see a lot more of the random juxtapositions, fandom content, chaos, humour, and weirdness that doesn’t make a Twitterbot feel that out of place. Perspective is key in how we interpret things, in life and online. Texts and images are shot off into the void and then appear to us, maybe through a tag, maybe a retweet, maybe a suggested tweet (ugh). It’s this fluctuating environment that made me laugh when the book quoted a tweet from a British Heart Foundation campaign: “You’re on Twitter. Your heart stops. You go into cardiac arrest. Like this tweet to find out what happens next.” I’m sure on a timeline filled with news and health information, this reads differently than it did to me, because to me an invitation to like a tweet and find out what happens after my heart stops sounds like someone's joke. All of this to say: I’m surprised the chaotic environment of Twitter did not feature more prominently in a book about Twitterbots.
Also surprising was the fact that the authors did not discuss current limitations when it comes to AI textual analysis and how a bot might make use of such tools. This was especially surprising when it comes to an environment like Twitter, which is full of slang, spelling mistakes, and people from around the world who speak (type) in different dialects. The authors seemed to gloss over these fuzzy areas, which I find too bad because to me they’re deeply interesting.
Additionally, I think it would have been interesting to compare/contrast bots made by programmers and academics with the bots I find I actually see most of the time online, which tend to be spambots that post suspicious links, advertisements, and propaganda.
Twitterbots was a useful book to read and gave me a new understanding for its subject and how they actually work, especially the more complex ones. Although I was at times frustrated with what it omitted, this book was able to serve as a jumping off point for me to think about those omissions, which I do appreciate. Finally, if you’re looking for a fun bot, check out AI Generated TV Tropes Names (@ tvtropes_gpt2).
My primary frustration with this book is that it sometimes ignores the environment of Twitter and instead comes close to treating it as nothing more than a platform the bots post on. This is strange to me, as the mash-up and juxtaposition function of Twitterbots is perfectly at home on Twitter itself and the internet more generally. The authors sometimes discuss this, particularly when it comes to how some users will use an opportunity to find a way to use hate speech or crude jokes when they interact with a bot or brand, and more generally in the last chapter. However, it sometimes felt like the authors forgot about it, particularly when discussing how Twitter bots can make the familiar feel unfamiliar via unexpected mash-ups or words or content, when this is in fact a huge part of the Twitter experience already.
Perhaps the authors’ take on Twitterbots is the result of the particular perspective of the Twitterbot community and those around it. The authors quote Erin McKean: “[Twitterbots] take you outside your news-outrage-and-sandwiches timeline to give you a minute to look at the world in a new way.” I’m sure a lot of people would not describe their timeline this way, and instead see a lot more of the random juxtapositions, fandom content, chaos, humour, and weirdness that doesn’t make a Twitterbot feel that out of place. Perspective is key in how we interpret things, in life and online. Texts and images are shot off into the void and then appear to us, maybe through a tag, maybe a retweet, maybe a suggested tweet (ugh). It’s this fluctuating environment that made me laugh when the book quoted a tweet from a British Heart Foundation campaign: “You’re on Twitter. Your heart stops. You go into cardiac arrest. Like this tweet to find out what happens next.” I’m sure on a timeline filled with news and health information, this reads differently than it did to me, because to me an invitation to like a tweet and find out what happens after my heart stops sounds like someone's joke. All of this to say: I’m surprised the chaotic environment of Twitter did not feature more prominently in a book about Twitterbots.
Also surprising was the fact that the authors did not discuss current limitations when it comes to AI textual analysis and how a bot might make use of such tools. This was especially surprising when it comes to an environment like Twitter, which is full of slang, spelling mistakes, and people from around the world who speak (type) in different dialects. The authors seemed to gloss over these fuzzy areas, which I find too bad because to me they’re deeply interesting.
Additionally, I think it would have been interesting to compare/contrast bots made by programmers and academics with the bots I find I actually see most of the time online, which tend to be spambots that post suspicious links, advertisements, and propaganda.
Twitterbots was a useful book to read and gave me a new understanding for its subject and how they actually work, especially the more complex ones. Although I was at times frustrated with what it omitted, this book was able to serve as a jumping off point for me to think about those omissions, which I do appreciate. Finally, if you’re looking for a fun bot, check out AI Generated TV Tropes Names (@ tvtropes_gpt2).
December 4, 2019
For a rich text version of this review and other writings, check out [DinaburgWrites.com]
Well.
This book was meant to be my victory lap: I began reading Twitterbots: Making Machines that Make Meaning while I was deep within a month-long interview to write for an AI chatbot language that I, somehow, already had experience writing in—kismet, found.
But in the end it was kismet, lost. Even before that option fluttered away on tiny bird wings, I began to feel trepidation toward Twitterbots: “Why am I reading about a subject I’m about to spend most of my time working on?” It went from interesting hobby to potential schoolwork. The constant refrain directing the reader to the authors’ web resources echoing from each page didn’t alleviate the scholastic feel. But those resources—equal part repetitive in mention as useful in practice—couldn’t break up the meaty parts of the text:
[ Read the rest of this review at DinaburgWrites.com ]
Well.
This book was meant to be my victory lap: I began reading Twitterbots: Making Machines that Make Meaning while I was deep within a month-long interview to write for an AI chatbot language that I, somehow, already had experience writing in—kismet, found.
But in the end it was kismet, lost. Even before that option fluttered away on tiny bird wings, I began to feel trepidation toward Twitterbots: “Why am I reading about a subject I’m about to spend most of my time working on?” It went from interesting hobby to potential schoolwork. The constant refrain directing the reader to the authors’ web resources echoing from each page didn’t alleviate the scholastic feel. But those resources—equal part repetitive in mention as useful in practice—couldn’t break up the meaty parts of the text:
Though human creators benefit from the same word-of-mouth marketing by avid followers, we don’t want our bots to simply ride on someone else’s coattails but to become an active part of the co-creation process.
Like Duchamp recognizing the aesthetic merits of a lowly object that many others have scorned, we become connoisseurs of the generative object trouvé when we acclaim these accidents of bot meaning.
[ Read the rest of this review at DinaburgWrites.com ]
March 26, 2020
This book is an academic treatise of the variety of bots roaming the Twittersphere. It successfully brings together the artistic endeavour and creativity required along with the coding methods needed to bring bots to life.
It takes you through the different type of bots through real examples, their authors, methods on writing your own along with insightful anecdotes and long, repetitive, philosophical musings.
I've always had an interest in bots, be it Twitterbots or other types, so this book served its purpose by teaching me more on Twitterbots to the point I feel comfortable trying to write my own. Links to web resources were offered, including a GitHub repo.
However, I found it a dense and verbose read due to the lengthy reflections of the authors and the many historical and artistic references. I had to persevere numerous times. It's not an out-and-out textbook for coders wishing to learn about writing bots. It's more suited as a book for artists looking to explore new avenues for their creativity.
It's certainly an interesting read if you're willing to persist. If you're more interested in the technical aspects of Twitterbots, then you are probably better served elsewhere.
It takes you through the different type of bots through real examples, their authors, methods on writing your own along with insightful anecdotes and long, repetitive, philosophical musings.
I've always had an interest in bots, be it Twitterbots or other types, so this book served its purpose by teaching me more on Twitterbots to the point I feel comfortable trying to write my own. Links to web resources were offered, including a GitHub repo.
However, I found it a dense and verbose read due to the lengthy reflections of the authors and the many historical and artistic references. I had to persevere numerous times. It's not an out-and-out textbook for coders wishing to learn about writing bots. It's more suited as a book for artists looking to explore new avenues for their creativity.
It's certainly an interesting read if you're willing to persist. If you're more interested in the technical aspects of Twitterbots, then you are probably better served elsewhere.
December 4, 2018
Very interesting but a complete pain in the ass to read. The author uses a dense writing style that feels way too much like a college textbook. A lighter touch would've made Twitterbots more accessible.
December 17, 2024
I gave this a low score, but I honestly think the book is pretty good, and I am just not the target audience. I learned quite a bit of vocab words on automated machines and some of the processes, but I wasn't in a position to recreate and explore the samples provided.
Displaying 1 - 7 of 7 reviews