Artificial Intelligence, or AI is a fascinating and powerful tool. There are virtually endless applications – chatbots, automated emails, and scheduling tools have revolutionised the small businesses that utilise them for marketing and the demand for this technology has created businesses that further innovate. Giant nerdsquee for AI.
Many things about AI, however, concern me (and they did even before I read Click Here to Kill Everybody by Bruce Schneier 5 years ago) so when I have conversations with clients about how AI could help them – it becomes clear that the risks of individual and collective adoption of these technologies are not widely known. So buckle in.
Throughout human history, we have continued to perpetuate a cycle of innovation leading to irreversible impacts on society. The earliest example of this was only a few hundred years ago and we have continued to make life more complicated and dangerous ever since.
The industrial revolution was supposed to reduce human labour, replacing it with machine labour. (Not for the benefit of the labourers who might then be able to live a life of leisure and peace, to allow their bosses to circumvent fair labour laws and continue to profit) The Luddites who opposed the introduction of technology in the textile industry for example feared that skilled craftspeople would be made redundant and understandably revolted. Because unfortunately, people who profit from technology don’t like to listen to scientists and philosophers. Or historians… they’re such party poopers.
Fast forward 212 years from the industrial revolution and we now have AI software scraping the internet or other data banks to generate images and text. AI has been used to write and illustrate books, compose music, produce poems, research, reports, articles, speeches, legislation…
There is no Global AI Governance to prevent or regulate this technology. There isn’t a machine we can smash (we’re not even allowed to peacefully protest any more!) and Pandora’s box has definitely already been opened.
Artists were not invited to consent for their work to be used or reproduced when art is synthesised by AI. The copy generated by tools like Chat GPT, Google Bard etc is derivative and filled with misinformation and bias. Given that humans seemingly cannot determine the difference between journalism and propaganda, evidence from deepfakes – you must concede that without regulation, AI is as dangerous as the internet is. The internet began in 1991 and we still haven’t created an adequate system to ensure the safety of its users. Whilst the various governing bodies around the world are theorising how governance could be approached, AI is becoming more widely adopted than people are aware of and is already having a negative impact.
The AI algorithms of social media that were created to help humans connect more led to global polarisation and extremism within about 5 years of being widely adopted. We still haven’t cracked online safety for social media or advertising. Our data is being mined by them all. We are the product, not the consumers.
Bias is a major issue when it comes to AI algorithms. AI is not neutral – it was created and coded and debugged by humans and humans have biases. Therefore AI learns the same classism, racism, ageism, sexism, ableism and other biases that humans do… and perpetuates them.
AI is used by education, healthcare, financial and government institutions to create models and automate processes. It creates inequality because it mirrors and exacerbates existing issues such as the gender pay gap or property redlining. Because these systems exist, AI upholds them. In the USA, AI already has the power to deny you a medical prodedure, insurance coverage, benefits, a mortgage, a university place, employment and even impacts the legal appeals process.
But wait, you were just thinking about using AI for your small business.
So yes, AI can create you a powerful logo. But it doesn’t know that you don’t want your logo to look like a swastika. Yes, AI can create artwork for your business. But it doesn’t care about intellectual property, leaving you vulnerable to legal issues. Yes, AI can write marketing copy for you. But it doesn’t have the 27 years of human expertise in doing so that I do in making that copy effective, accessible and relevant to your audience.
No matter how sophisticated a process you create, you cannot synthesise human artistry and expertise. For art to exist, a human has translated something they felt internally into something that exists externally – a song, a sculpture, a painting, a poem, a photograph… It will never truly pass the Turing test. Nor should it.
Butterflump 