I ALWAYS know there is something fishy going on with a new tech product when journalists start desperately reaching out to science fiction authors to explain it for them. Such is the case with ChatGPT, an artificial intelligence chatbot from San Francisco company OpenAI, which has become one of the world鈥檚 most widely used apps in just a few short months.
So many news outlets were asking science fiction writers to weigh in on AI鈥檚 capabilities that the Science Fiction and Fantasy Writers Association had to issue a on its website, linking to dozens of authors鈥� thoughts on the matter.
Advertisement
Typically, when a cutting-edge tech product comes out, you would expect engineers, scientists and researchers to comment on it. But OpenAI isn鈥檛 a typical company. It deals in myth-making and hype-spinning, and its representatives portray the firm鈥檚 products as , or human-equivalent consciousness. The OpenAI blog is packed with science fictional scenarios (many written by company co-founder Sam Altman) about how ChatGPT and the firm鈥檚 picture-creating product DALL-E are on the cusp of utterly transforming humanity 鈥� and possibly even destroying it.
I am not saying that ChatGPT isn鈥檛 a fun little app with lots of applications. But at this point, its notoriety comes largely from marketing. Altman and other AI entrepreneurs in Silicon Valley are fond of saying that AI products may cause because they could become superintelligent, conscious beings at any moment. And then they would have the power to wipe us all out, or , or something even more bizarre.
The question is, why would you use such a dark vision of your product to market it to people? Partly, it is to make a rather silly chatbot like ChatGPT sound a lot more formidable than it is. If you can be convinced that current AI apps will go Skynet any day now, then maybe you will buy the idea that ChatGPT is a powerful, revolutionary breakthrough.
The other reason to use this kind of dystopian hype is that it allows companies like OpenAI to cast themselves as the heroes that will save us from the coming chatbotpocalypse. Contributors to the OpenAI blog have spilled a lot of electrons in about how to bring AI values into alignment with human values. That way, when ChatGPT becomes a sentient superbeing, it will share our values and mindset, and so will ally itself with us instead of simply nuking us from orbit.
That sounds incredibly weird, until you realise that this is a marketing formula as old as advertising itself. Recently, I was discussing this very topic on a panel of science fiction writers at the Tucson Festival of Books in Arizona. Veronica Roth, author of Arch Conspirator and the Divergent series, pointed out that OpenAI鈥檚 approach reminded her of the diet industry.
To sell their get-slim-quick products, diet companies need to convince consumers that there is something wrong with their bodies 鈥� preferably something they had never noticed before. Ads bombard them with images of the 鈥渋deal鈥� legs, for example, and consumers go from never thinking about their ankles to agonising over all the fat on their fibula bones. Luckily, Diet Bomb is here with a special formula to melt the pudge off those offensive ankles!
Like Diet Bomb, OpenAI has invented a problem that it promises to solve 鈥� if you will pay, of course. Recently, the firm it would start charging for access to ChatGPT鈥檚 services, and struck a $10 billion deal with Microsoft to bring ChatGPT to the company鈥檚 Bing search engine.
OpenAI鈥檚 sales pitch is a brilliant rhetorical trick, and it works 鈥� until you recognise that you have fallen for one of the oldest tricks in Ye Olde Book of Hype.
The reason why media outlets are asking science fiction writers for their thoughts on AI apps isn鈥檛 because we are about to enter a wild new future world. It is because , such as University of Washington computational linguist Emily Bender, they say that ChatGPT is nowhere near being a human-equivalent intelligence. Artificial general intelligence is still very much in the realm of science fiction and, frankly, it may always be.
So, when you read about these amazing new chatbots that are about to write great novels, invent a cure for ageing and replace humanity with paper clips, just remember you are being presented with an ad campaign. The people who make chatbots want you to buy their products. They aren鈥檛 trying to elevate humanity or protect us from a dangerous, existential threat. They just want you to believe there is a problem out there that only AI companies can solve. For a price.
础苍苍补濒别别鈥檚听飞别别办
What I鈥檓 reading
The Best of all Possible Worlds by Karen Lord, a brilliant reimagining of the alien first contact story.
What I鈥檓 watching
Starstruck, the perfect romantic comedy for self-sabotaging nerds.
What I鈥檓 working on
Starting to write my fourth novel by researching Roman agriculture around Pompeii in the AD 50s.
