[..] language is so central to our understanding of each other that when we encounter language that doesn’t actually reflect the thoughts, ideas, or communicative intent of another person, it’s difficult not to imagine some humanlike mind behind it.
page 7, Chapter 1: An Introduction to AI Hype, The AI Con by Emily M. Bender & Alex Hanna
Table of Contents
Why The AI Con?
Anyone who follows me on social media and/or anyone who regularly reads this blog, you will already know I very much despise LLMs and the “AI” hype bubble we are currently living in. So when a friend recommended The AI Con to me recently after a discussion about our mutual dislike for all things “AI”, I immediately put it on my TBR and on my library holds list.
I have read a fair few “AI” polemics at this point, most of them long form blog posts, and I have written one myself so I consider myself somewhat of an aficionado of this specific sub-genre of technology polemic. My friend said that this book was a fairly easy read and contained some light humour in it which intrigued me more since I am not particularly in the mood to read a heavier book.
This is my second non-fiction read this year and it also happens to be the second dual author book I have reviewed for this blog, the first being This Is How You Lose the Time War. This is my first time reading anything by either Emily M. Bender or Alex Hanna so my only expectations came from what my friend told me about the book.
Let’s get into it.
The Book
The AI Con starts off with a preface introducing the authors and what the book is going to be, it is then divided up to seven chapters, each tackling and deconstructing a specific aspect of the “AI” hype bubble. Before I start talking about individual chapters I think right from the jump I appreciated the straightforward, concise prose of this book. The authors make their point directly and don’t beat around the bush. Here is a quote from the preface to demonstrate that.
We are writing now, in late 2024, from the inside of what feels like the height of the AI hype bubble. As we say on the podcast: each time we think we’ve reached peak AI hype–the summit of bullshit mountain–we discover there’s worse to come. We’re using what we see in this bubble to document the contours of hype about AI, its causes and its short- and long-term effects. Our primary goal is to inhibit the next tech bubble. We hope that by pulling back the curtain, we’ll help you to be able to spot the hype now and the next time around, while honing your own needles.
Preface, XI
One note about formatting – I’ll be breaking up some of the quotes from The AI Con below into paragraphs for the purposes of readability. This is the one major stylistic criticism I have of this book, the authors use overly long paragraphs throughout the text. Thankfully, its not too detrimental to the readability of the book since the prose is quite straightforward, but it was definitely a minor annoyance for me.
On “An Introduction to AI Hype”
This as the title suggests an introduction to the “AI” hype bubble. Most of this chapter is about things I already knew about so it i was not new information to me. But there were some bits of flavour that I didn’t know about like the Chuck Schumer bit or the whole p(doom) aspect of this bubble.
Artificial intelligence, if we’re being frank, is a con: a bill of goods you are being sold to line someone’s pockets. A few major well-placed players are poised to accumulate significant wealth by extracting value from other’s people creative work, personal data, or labor, and replacing quality services with artificial facsimiles. The language of p(doom) is a ruse to keep us focused on imaginary scenarios, filled with awe at modern robber barons’ allegedly potentially world-ending technology and too distracted to see the daily harms being done in its name.
page 4, Chapter 1: An Introduction to AI Hype
I think this chapter is a good introduction for anyone who has heard of LLMs but doesn’t know anything about the whole ecosystem of hype that surrounds it.
On “It’s Alive! The Hype of Thinking Machines”
This chapter deconstructs the myth that these large language models are “thinking” machines of any kind. I really appreciated the way the authors deconstructed this particular aspect of the bubble and it is one that really grinds my gears whenever I hear people consciously and unconsciously anthropomorphizing one of these LLM bots.
But to be clear: neither large language models nor anything else being sold as “AI” is conscious, sentient, or able to function as an independent, thinking entity. Despite the proclamations of corporate and academic boosters like Aguera y Arcas and Minsky, technologies that synthesize text or solve mathematical puzzles are not artificial life-forms. However, it serves many people to say so: entrepreneurs who have a product to sell, researchers who have academic departments to fund, and zealots who have institutions or followers that would benefit from such a fiction being perpetuated.
page 22, Chapter 2: It’s Alive! The Hype of Thinking Machines
I also learned some new things from this chapter, namely what the hell a neural network actually is and why it is in fact a very misleading term.
Neural nets are composites of mathematical functions called “perceptrons” that each take in multiple inputs and then run a calculation to determine what value to output, based on those inputs. The perceptrons are connected in a network, such that the output of each can serve as the input to many others and each of those connections is associated with a “weight”, which can be interpreted as the strength of one on the next perceptron in the network. “Neural network” is an impressive-sounding but very misleading term: they are named as such because perceptrons were very loosely inspired by a 1940s understanding of how neurons work in the human brain.
page 25, Chapter 2: It’s Alive! The Hype of Thinking Machines
The chapter ends with connecting the contemporary discussions of intelligence to race science.
Discussions of intelligence, pertaining to people or machines, are race science all the way down.
page 35, Chapter 2: It’s Alive! The Hype of Thinking Machines
On “Leisure for Me, Gig Work for Thee: AI Hype at Work”
This chapter connects the hype to the workplace and I found this particular quote about LLMs in relation to the craft of writing particularly striking.
There’s a few reasons you may want to resist temptation. First, for those of us for whom writing is a key part of our jobs, we would argue that the act of writing itself is co-constituent with thinking critically and creatively. This is an argument from craft: critical thought is co-created with creative expression, whether that is written, spoken, or signed speech, drawing, playing music, or physical movement.
For instance, qualitative sociologists are typically taught that their written memos–the writing they may undertake after reading through interview transcripts or ethnographic field notes–are really where their analysis occurs. Pedagogically, writing something in our own hand or in our own words encourages understanding, rather than rote memorization. Writing is intertwined with the act of thinking not separate from it.
page 50, Chapter 3: Leisure for Me, Gig Work for Thee: AI Hype at Work
As someone who writes both for their job and as a hobby, yep, 100% agree with all of this.
On “If it Quacks Like a Doc: AI Hype and Social Services”
This chapter is all about AI hype in the field of social services. To be quite frank, this is where the bubble gets really goddamned scary for me. The thought of a doctor using an LLM for the purposes of diagnosis instead of their own training & experience is terrifying. On the similar note, using one of these sycophantic compliment bots for the purposes of therapy is disgusting.
Government processes that affect people’s liberty, health, and livelihoods require human attention and accountability. People are far from perfect, subject to bias and exhaustion, frustration, and limited hours. However, shunting consequential tasks to black-box machines trained on always-biased historical data is not a viable solution for any kind of just and accountable outcome.
page 76, Chapter 4: If it Quacks Like a Doc: AI Hype and Social Services
On “Artifice or Intelligence? AI Hype in Art, Journalism, and Science”
This chapter goes further into breaking down the ways AI hype is permeating the arts, journalism, and science. Naturally, I found myself gravitating to the author’s thorough refutation of the promotion of AI art as a form of “democratizing” art. That line of argument is one that also really grates at me.
Lastly, the promotion of AI art betrays a deep misunderstanding of the nature of what ought to be considered art. The major functions of art include sharing experiences and providing insight into the human condition–not to mention the joy and fulfillment of artistic expression.
As philosopher and technology scholar Johnathan Flowers has said, the purpose of art is to signal a particular kind of intention and to convey a particular type of experience, and this is precisely what AI art lacks. Art exists as a means to convey something about the human condition.
A diversity of art forms exists because we, as humans are diverse. An expression of human experience can be simple or complex but need not involve a high level of technical acumen. By this measure, we find the claims that crafting prompts is akin to “democratizing” art and producing the same joy to be unconvincing.
page 109, Chapter 5: Artifice or Intelligence? AI Hype in Art, Journalism, and Science
On “I’m Sorry, Dave, I’m Afraid I Can’t Do That: AI Doomers, AI Boosters, and Why None of That Makes Sense”
This chapter gets into the weeds about the different kinds of AI bros out there and how both AI Doomers and AI Boosters are two sides of the same coin and are both equally nonsensical. I knew some of this but a lot of this detail was new information to me, especially the bit about “alignment”.
Embedded in the idea of alignment is a premise with we fundamentally disagree: that AI development is inevitable. We take umbrage with this on multiple grounds. For one thing, what is currently being developed as “AI” does not work, nor is it helpful, for an overwhelmingly large portion of people living on the earth today, especially people in the Majority World.
Furthermore, as we’ve said elsewhere, there is no clear, precise definition of AI. Nor is there any solid evidence that the work of AI research now (or of the past seventy years) is on a path towards that undefined destination.
Lastly, the development of mass automation tools is not socially desirable. If you’ve gotten this far in this book, you’ve seen how these technologies serve as a means of centralizing power, amassing data, generating profit, rather than providing technology that is socially beneficial.
In other words, this is a choice, one being made by powerful interests, but one that the rest of us do not have to go along with. Thus we reject this inevitability out of hand.
page 143, Chapter 6: I’m Sorry, Dave, I’m Afraid I Can’t Do That: AI Doomers, AI Boosters, and Why None of That Makes Sense
I also learned what the term TESCREAL actually stands for. It is a term I’ve heard before but never had it fully spelled out. For those curious: TESCREAL = Transhumanism, Extroprianism, Singularitianism, Cosmism, Rationalism, Effective Altruism and Longtermism. I knew only about two of these – Transhumanism and Effective Altruism.
On “Do You Believe in Hope After Hype?”
The AI Con uses its final chapter to talk about strategies one can use to combat AI hype in the various places that it has wormed itself into. It provides a set of very useful questions that one can ask to push back on the hype.
- What is being automated? What goes in, and what comes out?
- Can you connect the inputs to outputs?
- Are the systems being described as human?
- How is the system evaluated?
- Who benefits from technology, who is harmed, and what recourse do they have?
- How was the system developed? What are their labor and data practices?
I think this set of questions is the most valuable thing I have gotten from The AI Con. Sure learning various facts and tidbits about the AI bubble was useful knowledge but these questions provide a strong framework to push back not only this particular bubble but on any such bubbles that will happen in the future.
Concluding Thoughts
The AI Con is a good introductory text for anyone who wants to understand all the key aspects of the “AI” hype bubble we are all in right now. As someone who was already aware of a lot of what the book talks about, this book’s usefulness as a whole was reduced but it did flesh out my knowledge and closed off any gaps in my understanding of these stochastic parrots.
I am glad I read The AI Con and I will recommend it to anyone who’s looking for a generally accessible book on the topic of “AI”. I am also planning on checking out the Mystery AI Hype Theater 3000 podcast that Emily and Alex host because that is the kind of technology podcast I’m into nowadays.
That’s all from me, see y’all in the next one.
Thank you for reading this blog post. Please consider supporting my writing work directly through Stripe or Patreon. Additionally, please share this post on your socials and/or with your friends. It really helps! If you have any thoughts/feedback/thoughtful criticism to share, feel free to use the handy contact me page to send that to me.