If the metaverse already lost word of the year to Gobelin Mode in 2022, this year it is facing a new competition: artificial intelligence. Bill Gates recently pointed that AI was “just as important as laptops and the Internet.” Two months after its debut, ChatGPT has over 30 million users and receives approximately 5 million visits per day. According to New York Times journalist Kevin Roose, this is “the fastest software growth ever observed – Instagram reached 10 million users in one year”.
The not so young among us will remember windows’ famous cartoon helper Clippy. He’d often make unsolicited apparitions when using the Microsoft Office apps and offered to help us, a feat he’d rarely manage to achieve. If we’re bringing him back from the digital graveyard, it is because, before the announced revolution it rumored to be, ChatGPT is nothing more than a Clippy on steroids, on Google’s side as well as Microsoft’s.
THE AUGMENTED HUMAN
Every time AI comes to the mainstream, the end of the human era is proclaimed. However, each time, it is just making it stronger. No need to be a chess player to remember Gary Kasparov being outperformed by DeepBlue, IBM’s supercalculator, and how it spelled game over, humans had been outdone by the machine. Today, the same calculator can’t come close to the current chess world champion Magnus Carlsen. And if he did claim that “engines are stupid” earlier in his career, he has recently admitted that nothing was more powerful than a human “powered by engines.” Currently, AI really is Machine Learning, rather than actual intelligence, and has a long way to go before it can send us back to the concept of Singularity that James Cameron promoted in his cult movie Terminator – I’ll be back.
Though AI is still in its infancy, its impact is already felt. It does not need to be a revolution to have a positive impact on our lives. Indeed, when asked about its reason to exist, ChatGPT answers that it was precisely designed to “improve and enhance the human abilities and skills.”
CNRS researcher Laure Soulier compares it to Wikipedia “When it came out, a lot of people were talking about “the end of school”. And we learned to use these tools. I am confident when it comes to ChatGPT. We will learn to use them, so they can help us to work differently”. It is in fact through this prism that we should tackle the topic of AI: a tool at our disposal, and not instead, of humans.
Above everything, this technology brings a new form of raw material to feed our imagination. This comes down to the nature of the tools itself, from ChatGPT to DALL-E or even Stable Diffusion. These AI platforms, specialized in text, image or video don’t create anything, but rather generate. Each of them scans the web and its contents, open sourced and copyrighted alike … and then integrate them to their database; their memory, you could say. In response to the more or less original demands - prompts - of its users, the platform remixes existing content into completely relevant or entirely preposterous results. No matter the outcome, it has the advantage of opening the horizon of possibilities. These tools therefore have the potential to inspire and accelerate the production of creative work, but not at any moment does it become a substitute for creativity.
Beyond their promise, the way these tools function is already triggering fundamental ethical questions. Artists are standing up against the absorption of their works to feed these platforms databases. The same tools are of course coded by humans, often sharing the same social and geographical reality, therefore imprinting into it their opinions, biases, and prejudices, with sometimes disastrous outcomes. These questions are essential, and too often perceived as obstacles to innovation by their creators. Talking about AI ethics is a sure way to getting called “scared.” When Rabelais wrote “science without conscience is nothing but the soul’s ruin”, he was terrified. Sometimes, dwelling on a genre that has largely explored these ethical questions can save us some precious time. Fans of hard science-fiction will inevitably think of Isaac Asimov’s Foundations:
I. A robot cannot hurt a human, nor passively let this human be exposed to danger.
II. A robot must obey the orders given by humans, except when such orders are contradictory to the first law.
III. A robot must protect its existence, to the extent that it does not contradict the first or the second law.
Establishing these rules rooted in sensible common sense as soon as possible seems to be the wisest course of action.
When ChatGPT is asked to define itself, its response is quite eye-opening: “I am a linguistic model without personal thoughts or feelings. My goal is to offer precise and useful answers.” This honesty could almost be endearing, if it were not referring us back to the tool’s inherent limits and flaws. In the words of its creators: “ChatGPT’s format should allow it to answer questions in dialogue form, to admit its errors, to contest incorrect information and reject inappropriate answers.”
An ambitious blueprint for a first version to which we easily lend creative intentions, from writing poetry to song lyrics. This brings us to the legendary singer, author, and composer Nick Cave, who reacted to a song “written” in his style by ChatGPT. His sentence is irrevocable: “this song sucks, it is a grotesque imitation of what a human being is.” Indeed, according to him, the creative act is not a matter of imitation, quite the opposite. It is through the confrontation of one’s own vulnerability, faults, and limits, facing what goes beyond us and tears the heart apart, that artists find what was lurking beyond the forefront of their minds. That is precisely why it is a profoundly human process, simply because algorithms don’t feel, because data does not suffer (except in the hands of some planners). ChatGPT’s tragedy is that it is destined to remix without ever being able to create a piece of work deserving to be called authentic.
- For 2022’s World Mental Health Day, health insurance company Helan used AI to create See My Pain, a campaign destined to make mental illnesses more tangible and easier to talk about.
The computer, the Internet, the iPhone, all revolutionary technological progress at their time, are today taken for granted. In several instances, Steve Jobs observed that what distinguishes the human species from the other stems from our ability to build tools. He specifically quotes a study comparing our speed compared to those of other animals showing that, unsurprisingly, we were at the bottom of the list. However, it’s another story when the same human is on a bike. This brings him to describe the computer as a “bicycle for the mind.” If a computer can serve as a bike, to what could we compare AI? That is the question we asked ChatGPT. His answer: “a toolbox for the mind.”
