Is AI unethical?

With Artificial Intelligence dominating headlines, its role in society is becoming increasingly hard to ignore. There’s no doubt that it’s yielding revolutionary advances for the better, in the medical field and in fraud detection for instance. But having also seen the negative implications, should we scale it back? At what cost do we allow its influence to spread and expand, and what sort of net result can we expect to see in a generation’s time?

The threat to creative industries

The 2023 Writers Guild of America strike centred largely around concerns over ‘generative AI’. Media companies are conducting various experiments with this to produce text content which, if successful, could put writers’ livelihoods in jeopardy. Not only is this a devastating prospect for creatives on a personal level, but begs challenging questions:

Can AI authentically replicate humans? It doesn’t encompass emotional intelligence, and I’d argue that as long as this is the case, its creations can never be classified as art. And when has a replication of anything ever measured up to the real thing?

What does art mean to us? Even if it could match humans, what would developing a complete indifference to the notion of AI generated art superseding traditional art say about us as a society? It’s one thing for Alexa to play Magic Radio on command and it’s another for organic emotions, experiences and expression to become indistinguishable from a computer.

Of course, some may argue that in content creation it’s simply a tool devised to churn out more material – a business strategy – and that aside from job loss for a small fraction of the population, there is no great disadvantage.

I would argue that while on a minor level it might not seem a terrible thing, devaluing human-made art could be the first step towards devaluing humans. After all, through all of history’s monstrosities, through all the government’s failings, through all our trials and tribulations, creating and engaging with art is what unites us, comforts us, stabilizes us, keeps us sane. Art is something that we, not computers, have always been able to take great pride in. Losing that pride, the thing that has separated us from technology, could exacerbate the distress felt by so many in these uncertain times.

While it’s not technically cheating (provided there’s some sort of disclaimer pertaining to the use of AI) it certainly feels like the gift and wonder of art, and what it means to be human, is being cheapened.

Artistic identity theft

However, author Joyce Maynard recently discovered that she had been deceived on a monumental level. An article in The Atlantic revealed that her work had been illegally scanned in order to feed an AI database in such a way that her novels could be replicated without her permission. She is taking legal action through The Authors’ Guild, but is there a loophole somewhere that will not only allow those behind the database to get away scot-free, but for this practice to take off and, again, threaten the livelihoods of real writers? Unfortunately, it seems Sunak’s suggestion that artists should re-train and get new jobs may in fact become necessary. What power do we have as humans to stop or even control such a powerful force?

Scams and agendas

Many employers now feel the need to state within job adverts that the use of aids such as Chat GPT are not permitted within applications. While they often state that the use of these aids is easy to spot due to AI detection technology, the effectiveness of this technology varies, and by contrast AI technologies are becoming increasingly sophisticated. I’ve heard numerous first hand accounts of people duping others through Chat GPT and getting high essay marks, and the news has seen its fair share of deepfake stories.

Deepfakes are synthetic media that, through digital manipulation, adopt the likeness of an individual. Unsurprisingly, despite the technology being in its infancy, we’ve already seen several examples of individuals being portrayed as saying, promoting or doing things that don’t represent them in the form of fake audio, photos or videos. In July, financial journalist and broadcaster Martin Lewis fell victim to this in an advert promoting a fake ‘Elon Musk investment’. Not only is the investment a scam, but Lewis was portrayed as its champion, making the incident a form of defamation. In May, a man from Long Island was arrested for superimposing the faces of women he knew onto images of other women’s naked bodies, and posting them online. Sadly, at the time of his arrest there was no law in place in New York State pertaining to this form of deepfaking, despite there having been a similar incident four years ago. He got six months.

So where do we go from here as a society? Does AI have the potential for good? – of course, and we’ve seen it. But I, and I’m sure many others, would argue that the potential for harm is more relevant and perhaps far greater. We may see many more writers’ strikes. A few days ago this year’s strikes came to an end following an agreement to draw up a three-year contract that will include protection against AI that was non-existent before. There’s no denying that this is a victory. Let’s hope writers the world over – not just in Hollywood – don’t have to continuously resort to strikes the way NHS staff, teachers and rail workers have in the UK due to unfair treatment. Let’s hope we can collectively step back and see AI for what it is – an innovative tool to be monitored, modified and restrained, not one to be glorified.