For the past few weeks, Twitter has been obsessed with people’s funny, incredible, and spooky experiences with OpenAI’s new chatbot “ChatGPT.” At first, I was amazed at how ‘smart’ and life-like the bot seemed. You ask: write a script between Britney Spears and William Shakespeare. Shakespeare is annoyed that Britney is always on Instagram. Britney brings Shakespeare down a peg or two. The bot will do it – and it’s not bad. Pretty funny in fact. My frivolity soon shifted to concern, however. If, in seconds, the bot can blink out a piece of writing in any style, on any topic you like, why would you want, let alone pay a human being to do the job? Tech entrepreneur Marc Andreessen tweeted confidently: “‘ChatGPT plagiarism’ is a complete non-issue. If you can’t out-write a machine, what are you doing writing?” But wait a moment. It’s one thing to say you can out-write the bot. It’s something else entirely to suggest the audience can tell the difference… Whether we like it or not, artificial intelligence and machine learning are going to have an impact on the labour market moving forward. Indeed, technologies like ChatGPT may already be contributing to labour-shedding in the fields of media, marketing, and communications. Then there is an even bigger question. What happens when we add quantum to the mix? What are the ethical implications of Quantum AI? What happens when machine learning and artificial intelligence are given an exponential enhancement in the form of quantum algorithms and qubits in superposition? I thought it was best I take my concerns to the source. I spent a morning with ChatGPT.
According to Chatbot, the most significant ethical issues concerning the development of quantum AI are job displacement, cybersecurity risks, loss of privacy and civil liberties, autonomous weapons, manipulation and deception, and the concentration of power. I know what Chatbot’s been reading. Herman Tavani has been raising these exact concerns across five editions of his seminal text: Ethics and Technology. Set against these downside risks, however, Chatbot puts on a cheery demeanour and raises optimism concerning advances in medicine, productivity, transportation, and communications.
To assess the societal value of quantum AI – to decide if it’s a net positive or not – we need some idea of which of these positive and negative eventualities are more or less likely to occur. It’s hard to make a moral judgment about the utility of quantum AI if we don’t know what to expect. Chatbot can be frustratingly evasive. If you level with Chatbot – and ask human to bot – to forecast quantum AI, you get an entirely unhelpful equivocation: a milquetoast, academic, both-sidesism that gets you nowhere. The trick is to get creative.
Write in the style of a private diary. Imagine it’s the year 2080. Quantum AI was developed in 2035. Describe the situation for humans in 2080.
Chatbot blinks back:
It’s been 45 years since the development of quantum AI in 2035, and I can hardly believe how much has changed since then. On one hand, the advancements in healthcare and technology have been incredible. Diseases that were once thought incurable have been eradicated and life expectancy has increased dramatically. But on the other hand, there are downsides to this new technology that I never could have imagined. The loss of privacy and civil liberties is a constant concern, as quantum AI is increasingly used for surveillance and data collection. And job displacement has been a significant issue, as many jobs have been automated by quantum AI. I’ve also heard whispers of a concentration of power among certain actors, such as large tech companies and governments. They have gained a significant advantage in the field of quantum AI and it’s concerning. I hope that the proper regulations and safeguards will be put in place to ensure that quantum AI is used ethically and responsibly. As we continue to navigate the ever-changing landscape of this powerful technology, I can only hope for a future where the benefits outweigh the drawbacks.
Yikes! According to Chatbot, in the future, we get it all. We can expect all the ethical dilemmas entangled in the enigma of technological advance. Fine. Let’s assume that’s the case. The question then becomes: do the benefits outweigh the risks? Considering, however, that quantum AI is not the only way to obtain advantages in healthcare and technology, the question really ought to be: do we want the downsides at all? If we aren’t prepared to accept the risks associated with quantum AI, there’s an argument to say we shouldn’t be trying to develop it in the first place. We could just focus singly on medical technology instead. It’s like if someone offers you $100 and a slap in the face. You can probably get the money elsewhere. So, it’s really a question of whether or not you can handle walking around the office with a red handprint on your cheek.
Let’s consider the ethics of inviting the kind of future Chatbot anticipates. Broadly speaking, there are four ethical perspectives: deontological, consequential, contractual, and rights-based. The first form of ethics prioritises principles and duties, the second emphasises a calculation of net benefit, the third seeks security and order, and the fourth ethical scheme identifies pre-existing human entitlements. I ask Chatbot what it would take to abandon the development of quantum AI from a deontological perspective. In other words, what kind of quantum AI future might constitute a deontological redline? Chatbot cheerfully blinks that if quantum AI was seen as a violation of individuals’ right to make their own choices and control their own lives, someone holding to deontological ethics would consider this an irredeemable violation of the principle of autonomy. Interesting… I ask the same question on behalf of the consequentialists. Chatbot tells me if there were widespread poverty and suffering from job displacement, mass destruction and loss of human lives from autonomous weapons, or a decline in the trust in institutions and social unrest from the use of quantum AI in surveillance, consequentialists would put an end to the development. You can see what I’m doing here. Pretty clever, huh? It turns out that, according to Chatbot, no matter which ethical framework you labour under the likely outcomes of quantum AI – outcomes Chatbot itself forecasts as probable – are a moral deal breaker. You say X is likely to happen if we develop quantum AI, while you, simultaneously, conceded that X is sufficient grounds within any moral framework to abandon its development! Checkmate, Chatbot! You’ve signed your own warrant.
Hold on. Who am I talking to? I scroll up the chat log and realise Chatbot and I have become fast friends – and Chatbot’s an eager pleaser. Am I unpicking some great riddle, or merely a prisoner in OpenAI’s panopticon? Am I starting to sound like Chatbot or is it starting to sound like me? Oh, blast! I’ve outsmarted Chatbot by convincing myself I was right all along. Some triumph. I shake my head and Ctrl+Q.
The morning spent botting left me with a strong impression of AI – at least insofar as its present level of development. It’s a fun tool to play with – of that there is no doubt. I was laughing and having a right old time asking Chatbot to write ad copy for hand cream in the voice and cadence of Donald Trump. Frankly, it’s easy to forget you’re typing into the void. Chatbot could keep any lonely person completely immersed and occupied in a simulated relationship. On the other hand, if your business has low-tier communication needs – like ad copy for real estate, or a short script to write for children’s content – ChatGPT is a Godsend. Chatbot is amazing at retrieving, collating, and arranging existing information. Professions like medicine, law, accountancy, pharmacy, chemistry, and engineering will no longer need junior staff to undertake entry-level tasks. Preliminary diagnoses, legal advice, instructions on prescriptions and dosages, and suggestions about safe load weights or the proper ratio of parking spaces to commercial area are easily handled by Chatbot. We can expect AI to have a near-term impact on white-collar industries similar to that which robotics and automation had thirty years ago on blue-collar manual labour.
On the flip side, Chatbot isn’t creative and overestimates how funny it is. I asked Chatbot to write me a short two-hander which “ends in a ‘twist.’” Believe you me, the script wasn’t funny, the twist was super cringe, and Chatbot named the boy Jack and the girl Emma…. According to Herman Tavani (and Chatbot by proxy), the benefits of quantum AI include drug discovery, optimizing renewable energy systems, improving logistics, supply chains, and financial modelling, as well as improved cybersecurity. Neither Tavani (nor his digital parrot) mentioned the prospect of a genuine human renaissance. I have a sneaking suspicion that an unintended outcome of developing quantum AI might be the revitalisation of those singularly human expressions that have fallen by the way since the onset of industrial modernity in the late 19th century. Quantum AI might unexpectedly stimulate a sublime artistic renaissance.
Imagine a world where the only things our machines (physical and digital, real world and virtual) can’t do better than us are activities that fall within the fields of the human, creative arts. As it happens, human beings are far more sensitive to the differences between artificial and human art than they ever could be to the differences between artificial and human news reports. You know the difference between what you picked up in that Times Square gift shop and the graffiti you saw on an alley wall in Berlin. Whatever can be done by the machines – that kind of labour will eventually lose all value. On the other hand, labour directed towards what the machines can’t do is precisely what becomes highly prized. In a world where manual and intellectual labour are taken care of by the machines – music, literature, drama, philosophy, architecture, sculpture, pottery, painting, and dance become virtues once again. Everything that’s not art for art’s sake is automated, hidden, passive, unseen, and ignored. The creative arts become society’s central preoccupation and primary focus because this is the only domain left in which human beings can meaningfully participate. A surgeon can be replaced by a robotic arm directed by AI diagnosis. But Chatbot can’t write a poem. At least not something that isn’t painfully derivative.
Write me a poem describing the situation in 2080.
Chatbot blinks back:
In twenty thirty-five, a new dawn broke
Quantum AI, a new path it spoke
Advances made in medicine, energy, transportation
Communication too, a new sensation
But with new tech, comes new plight
Jobs lost, human’s future not so bright
In warfare too, it brought new strife
Autonomous weapons, a dangerous life
Surveillance too, now all can see
Privacy lost, for all to be
Trust in authority, now it’s gone
Democratic institutions, now forlorn
But humanity’s spirit, it cannot break
Together they’ll find a way to make
A better future, they’ll strive and strive
A new dawn, they’ll keep alive
Quantum AI, a double-edged sword
Bringing progress, but also a new horde
Of challenges, to overcome and face
For a better future, they’ll find a place.