Generative AI should be used to transform society, not put dogs in Van Gogh paintings
As a technology journalist, I am open to technological changes and their impact. But the recent rise of generative AI has left me lacking enthusiasm.
This doesn’t necessarily have to do with the existential threat that generative AI may have on “content” – a large part of my work – and various jobs, art forms, decimation of information and more. Instead, my dim view of AI stems from a sense that AI should be further pushed to augment human experiences and society, rather than simply taking on work that is typically human-centric.
Don’t get me wrong: Generative AI can do some very useful things; I’m impressed with the AI-powered transcription and live translation. I also appreciate Google’s Magic Eraser when it comes to removing unwanted objects from photos. And AI can help with daily organization.
Additionally, Google’s NotebookLM can act as a form of educational tool by creating podcasts from the articles it receives; an evolution of this, said to be called Listen dailycan turn articles in the Google Discover section of Chrome and Android phones into podcasts; this sounds really useful.
However, almost every ad or use case I’ve seen for generative AI stems from making mundane human tasks easier, and often in ways that I find to be at best mundane and at worst completely joyless.
Sure, Generative Edit via Galaxy AI on the Samsung Galaxy S24 can completely rework a photo, but where’s the thrill of going out and creatively creating the photo you want from scratch?
Apple’s recent ads for Apple Intelligence show how generative AI can turn an informal, poorly written email into a formal letter. But forgive me for sounding a bit elitist: if you’re a professional who can’t write a good email, then maybe you should consider some tutoring instead of relying on AI.
The same goes for works of art; Sure, generative AI can help rework a few art objects for games, for example, but if it all works out, what’s the real artistic merit? AI art is also a painful middle ground.
I’ve also received ads touting how generative AI can help marketers better create content and generate leads, which is essentially a tool for an, arguably, supporting business function. It’s all so boring. Where are the AI systems that will actually revolutionize things for the greater good, rather than help speed up the creation of spreadsheets or flyers?
This push to make human-led tasks easier also has a sting in the tail, with stories of students using ChatGPT to write their essays, people using it to apply for jobs en masse, and other ways to produce content potentially without any real thought process and at the risk of eroding education and understanding.
Do not do as man does
The gist of my argument is that I don’t think AI should try to do jobs for humans. Sure, it might give companies in a capitalist world a little more money to tout in their annual financial reports, but at the expense of price. of a creative and vibrant people-led labor market.
I don’t want to sound too socialist, but ultimately corporations exist as a mechanism to help society function so that we don’t all beat each other’s heads up for a loaf of bread. If AI replaces a lot of human work, where does that leave society? Generative AI could give humans the space to do different things, but its functions so far seem to be performing human-like tasks at scale rather than extending them, and this makes me wonder what future jobs AI will unlock in instead of replacing.
I don’t think AI should do art, poetry or other creative pursuits. Ultimately, generative AI simply produces an amalgamation of what we already have, based on the way large language models are trained. It doesn’t really create anything truly new the way people do; Admittedly, there is a whole discussion to be had here about what is really new – see Alan Turing’s article “Computing machines and intelligence” – but that’s for another time.
My point here is that generative AI is seemingly being created to take on tasks that humans can already do; may not always be easy, but with some training and effort, virtually everything these AIs can do is within the reach of bipedal organic meatbags.
It’s the same idea as generative AI systems that can seemingly beat the Turing Test; they are programmed to mimic humans rather than being their own thing.
And it all seems so painfully uninspired, despite the technical prowess to get to this point.
Think differently
Instead, I would much rather see generative AI used in ways that actually enhance human lives and society as a whole by supporting them in tasks and functions rather than replacing them.
Recent reports in the US and Britain indicate that there is a shortage of qualified teachers to fill vacancies in the education systems. To compensate for that shortfall, I could use generative AI as a teaching support tool, by letting students ask questions to a smart system while their teachers are busy, or to organize additional lessons. Generative AI could also be used to support people with learning disabilities 24 hours a day or provide a framework for homeschooling for children who live remotely or struggle with a traditional school environment.
At the other end of the spectrum, with a growing elderly population, there is a care burden on families and society that has not been as acute as before, especially in Western families where families tend to be more dispersed. As such, healthcare delivery, both state and private, has become more challenging.
Here I could imagine a generative AI-powered system, where natural language understanding acts as a form of 24-hour support, helping answer questions from the elderly and – with the help of robotics – completing tasks such as running errands for them when a The human caregiver may be unavailable or overloaded with other tasks.
The same could be true in agriculture, where AI-powered robots could fill a shortage of farm workers. Or the robots could help farm workers, for example by feeding livestock or stacking hay bales, monitoring the lambing season overnight and supporting human workers rather than just performing a task for them.
There is also room for generative AI to be increasingly used to help people who are seeking therapy but may not feel comfortable opening up to a human; this would not replace the function of human therapists, but would increase access to therapy for a broader population.
Look, I’m not against generative AI in its current form. It has numerous shortcomings, problems and negative environmental impacts, all of which need to be addressed for a safe AI future. Yet its potential is enormous, ranging from a useful tool to potentially transformative technology.
It is this last point that I would like to explore further. Essentially, I’d like to see AI used to address problems in society, not just helping Joe Idle write an email or Sandra Snappy edit a photo rather than learning proper photo editing.
A lot of research is already being done into AI technology, but we are far from a utopia of smart technology; heck, self-driving car systems, which are almost certainly safer than the majority of human drivers, are not commonplace despite the use of advanced technology.
Ultimately, I think tech giants like Apple, Amazon, Microsoft, and Google could leverage their technology into generative AI designed to benefit society, rather than produce punitive profits and undermine education. We could see AI that could actually make the world a better place for the benefit of, not as a replacement for, people.