Generative AI and ChatGPT – you knew I’d get to it eventually… I am not an early adopter. I won’t rush out and grab the new thing because I prefer to sit back, watch how it evolves, see the bugs get worked out … that sort of thing. And if something is going to fall over I don’t want it taking me with it.
We’ve been surrounded by AI for a while now: self-serve checkouts are an early AI; Facebook and other social media algorithms serving us only what we want to see; our smart phones always listening; Dragon and other voice to text technology; Google Home, Siri, and Alexa. Generative AI – large language generative AI – is different and has pros and cons that I won’t go into in detail here. For now, my observations are that large language generative AI like ChatGPT is being pushed either by developers (understandably) or those early adopters – before they’ve really had a chance to prove its capability – or assess its dangers. And a lot of those early adopters seem to also fall into the ‘inauthentic’ camp.
Authenticity is something I am passionate about – it’s in my branding!! I am not on the ‘fake it till you make it’ bandwagon. Authenticity is absolutely essential in the virtual/remote space.
In the interests of authenticity, yes I’ve given ChatGPT a go and whilst I can see some definite uses in things like helping generate blog post headings (it did this one), keywords, SEO-related work, alt text – what I found was even in those tasks, the AI came up with the same words and style of writing – and definitely needed tweaking from me to make it sound ‘more human’ and less ‘wooden’. I’ve also trialled AI with voice to text – there’s no point me advising clients on its application if I don’t know the capability. I can tell you there is no way you would let the document produced by it go out as is. A human being with the requisite skills gives you a far richer and accurate result, and whilst it may not take 30 seconds, the result should require little if any post-production effort. Why? Because human listeners are human speakers which makes the human ear and perceptual system uniquely equipped to understand speech.
Insight – AI: Friend or Foe
I will post more on ChatGPT in particular, and AI generally, later, but for now I wanted to comment on an episode of Insight on SBS that I watched recently – and that really made me quite miserable – as a business owner … and a human being. Whilst the show seemed heavily skewed in favour of participants with positive stories, like medical professionals using AI to help diagnose disease or in the area of fertility, or robotics engineers who are doing great things in agriculture, there were a couple of participants who were really concerning. I won’t talk about the mental health implications of some of the commentary, but focus here on the implications of the tech from a business perspective.
One such user, small business owner Marina Pullin from South Australia who runs Jungal, proudly said she managed to get over the pangs of conscience she had when implementing ChatGPT. At the time her business – ironically a network of freelancing professionals – was expanding and she had need of two junior marketing roles, having already filled the marketing lead role. But putting generative AI to the test, she found she could not only do away with having to fill the two junior roles – but she ditched the marketing lead staffer as well. That’s three people out of a job.
Her justification was the “2,000% increase in engagement” – she doesn’t say what that equates to in sales but the word ‘engagement’ refers specifically to social media – where she is doing most, if not all, of her marketing. Social media engagement doesn’t always equal an upswing in sales. She does give away however, that it’s actually about bottom line. Doesn’t matter that three people are out of a job – “it’s all about economics” and apparently “We’re driven in business on short-termism”. That’s an interesting way to view business – politics maybe, but business?
“Short-Termism (noun): Concentration on short-term projects or objectives for immediate profit at the expense of long-term security.”
Oxford Dictionary
The use of the phrase ‘short-termism’ really concerned me. I’m not sure that’s what drives my business. Adelaide business consultant and financial concierge Evan Bulmer agrees. “You only exist in business because you have a social licence to,” he told me when I mentioned the ‘short-termism’ idea. In other words, people use your product or service which allows you to stay in business. If that need or want for your offering dries up, your ‘licence’ expires.
I’ve posted before about discovering your business purpose – and it probably bears repeating the quote from Seth Godin used in that article:
“… when the only rudder you have is ‘profit now,’ expect that your long term prospects are in doubt…”
Seth Godin
A quick look at the Jungal website and here is what annoys me the most: inauthenticity. The About Us page proudly uses the word “human” – and highlights it – together with “person” and “people” multiple times. Apparently it’s “Free and easy to be a great working human” – except if you want to work in marketing for them apparently. Further on: “We don’t just look after freelancers we look after humans” (except those working in marketing for Jungal) and proudly boasts “Transparency, Integrity, Community”. Interesting words ….
Transparency – it’s not transparent from the site that AI looks after the marketing for the business; we can assume it’s also used to generate the page content. The constant use of the words ‘human’ and ‘people’ smacks of it ‘doth protest too much’.
Integrity – well I guess we’d have to talk about that with the marketing lead now looking for work.
Community – I guess if you’re a freelancer looking for a network to join then that qualifies as ‘community’.
Later in the same episode of Insight we learn of Erin Arnott – a copywriter who was recently ‘let go’ (that’s a euphemism for ‘sacked’ by the way) from her job. Whilst her employer didn’t specifically mention AI replaced her, and there may well have been other considerations, she “has her suspicions” because in the weeks leading up to her dismissal she wasn’t writing blog posts anymore from scratch – she was tweaking ChatGPT generated content.
The participants agreed that some jobs will go but others will emerge – like robotics engineers. How does a copywriter pivot to become a robotics engineer – without the outlay of thousands to retrain, from a position of unemployment?
That didn’t seem to matter much to those in the audience who simply reiterated “Oh well, some jobs will go and others will emerge.” With global unemployment running quite high, can we afford to be quite so flippant or ‘uncaring’ about the employment status of our fellow humans?
Why should I care if you don’t have a job?
Here’s what I think we are missing: as with the rise of ‘cheap labour’ in offshore countries that were taken up in droves by business owners wanting to save a buck that sat like a dark cloud over the virtual assistant industry for a while from around 2007, this is precisely what seems to be the driver behind the take up of generative AI like ChatGPT: bottom line. Save on staff costs by having AI do the work for you! But that has wider implications. The rush for profit is leading to irresponsible behaviour – both by tech companies and those using it.
The Writers Guild of America – joined by SAG-AFTRA – is currently on strike, primarily over pay and conditions but they are also wanting regulation over AI creating scripts. The irony is AI could potentially put them out of work while simultaneously stealing from their work – since it is trained on data already created. This is why the WGA proposed in May when they attempted some kind of regulation to protect their industry: “AI can’t write or rewrite literary material; can’t be used as source material; and MBA-covered material” – in other words their material can’t be used to train AI. This was rejected by the Alliance of Motion Picture & Television Producers (AMPTP). As for the actors, in a shocking example of life imitating art, the AMPTP apparently offered “a groundbreaking AI proposal” that would require actors’ consent to create AI replicas – can anyone say Black Mirror?
Back here at home, it’s not speculative. Those four local jobs mentioned above are gone – four people out of work; four people not paying tax and contributing to society by having the funds to participate in it. What does that say about us as a society as a whole? It’s easy to say well, poor them – they should’ve thought about a different career path; or they can retrain; or the throw away ‘some jobs will go sure but others will emerge’ – so too bad if you picked the wrong skill set? How do you train for jobs we can’t even envisage existing yet? As we can see from the above, there won’t be a sector not touched by AI.
Unemployment and mental health
Having people employed is good for everyone. If it’s really “all about economics”, a basic understanding of economics tells us if people aren’t paying tax in the country they work in there is no money for services, hospitals and healthcare, schools, roads and other infrastructure, welfare …. There is no money being spent in local businesses.
It is well known that not having a job can affect mental health and impact physical health and relationships. Having a job gives you a sense of security and purpose, you have a certain ‘status’ in community, you can engage actively in society, and have a sense of accomplishment. In short, people are better when they are actively employed and engaged in work they enjoy and find fulfilling. The work you do is part of your identity and losing your job can result in a feeling of not being worthy and being isolated.
Post-COVID we are looking at high rates of unemployment globally. With increasing uptake of generative AI like ChatGPT and business owners being sold the notion it can replace various roles within their organisation, I do feel we are at risk of losing something more significant than jobs.
One ray of hope during the Insight program came from Professor Toby Walsh, Chief scientist at UNSW’s AI Institute who said: “Technology is not destiny. We get to choose where it goes. We need to work out how we ensure we include everyone. It has great scope for good, but it won’t happen unless we make the right choices.”
I’ll post more on this and some of my concerns about AI in future articles. Stay tuned!
References:
- SBS News: Marina Replaced Her Staff with AI – This is how it’s disrupting every day Australian jobs
- The Ringer: Hollywood’s Double-Edged Reckoning with AI
- National Library of Medicine: Effects of unemployment on mental and physical health