​Matt Flenley, Head of Marketing at Datactics explains how ChatGPT can affect the future workforce – and how you can use it to your advantage!

Having worked in the award-winning company Datactics for over five years, Matt Flenley has a wealth of experience in data and AI within the fintech industry and beyond.

Today he gives VANRATH his insights into the world of ChatGPT – the new computer programming technology that has taken the internet by storm – and explains how you as an individual or business can use this as an opportunity to thrive and get ahead.

Back in 2017, AI grandee Manoj Saxena said that AI will “amplify the knowledge of existing workers,” augmenting human decisions with computer power. Saxena should know, too: he was the first general manager of IBM Watson between 2006 and 2014.

What is ChatGPT and how does it work?

More recently, ChatGPT has hit the headlines, as people ask it to create blogs, conduct SWOT analyses on their competition, or suggest recipes for what’s in their cupboards (in rhyme, too, if that’s what you fancy!). It has plenty of promoters hailing the speed of content generation, and just as many detractors bemoaning the lack of variation and invention that one AI implementation can offer.

It’s broadly true, too: if you ask ChatGPT similar questions, it’ll give you similar answers, which could lead to a lot of essays, blogs, articles and indeed Tuesday dinners to eventually merge into one homogeneous blob.

This is because all machine learning requires a training dataset from which it can learn, thus that dataset is going to be made up of real material already in existence. AI art such as that created by Midjourney generates one-off images that people hail as wonderful for having no issues of copyright, except… the AI could only generate such imagery because of how it had been trained.

It needs to have seen the Mona Lisa to be able to create similar art.

What does this mean for the future?

Now it could be argued that this is just the next step in human development, a bit like saying without Handel there is no Haydn, or without Billie Holiday there is no Lorde.

On the other side of the fence would be people saying that art movements formed as revolts against the status quo, such as cubism emerging as a fight against why we must always have a literal representation of people, places and things.

In that light, could AI truly invent something new?

Discussions like this can make AI seem abstract. But in reality, it’s already embedded in everyday life. It’s been designed to work that way, recommending some Netflix films here, a Spotify track there, or your likely fastest route home from work on a Friday. The truth is that the machine learning augmentation of life is already here, and it just depends on how aware we are of the possibilities and risks for the world around us.

Back to Saxena. I was lucky enough to hear him speak at an AI event in London in 2019, where he was also asked the question of whether AI really will take your job or not. His response was succinct: “AI isn’t going to take your job any time soon, but someone who can use AI will.”

How can you use ChatGPT to your advantage?

So how do we become AI-augmented in our daily lives? How do we become a machine-learning native workforce? I’d recommend looking at three things.

1. Firstly, do your research. AI can augment your capabilities if you use it in the right way: leverage the massive datasets it’s built on, test several, and don’t just take verbatim the output of one model.

We do this a lot at Datactics, augmenting data matching using deterministic rules (that is, something identical, or close, or sounding or meaning the same thing) with data matching that employs AI techniques such as ‘K nearest neighbour’ or ‘Random Forest.’

In a ChatGPT sense, I’d recommend you ask it better questions about how you’d structure a blog, or maybe explore the threats facing brands like yours in the marketplace. Because it can be so ubiquitous in everyday life, we can be unaware of the biases that are baked into a training model, either intentionally or through accident. To overcome bias, for example if you’re conducting research on a market, maybe ask someone from a different socio-economic background than you would find most useful.

And then augment that using human desktop research, human heavy lifting, and scientific methods of testing and validating outputs.

2. Secondly, your humanity is your originality. You will have generated a lot of outputs using different models, human research, study and crowd-sourcing information. Now, sit atop it, and find something new, interesting or quirky. The way you present and communicate information to your audience will rely as much on your ability to deliver it in your own inimitable fashion as it will on the content you present.

Keep thinking back to Saxena’s statement that AI can amplify your knowledge. Make the best sound that sounds most like you, and amplify that.

3. Lastly, remain curious. AI models are extremely good at discovering and surfacing information they’ve been asked for, and perhaps they’ve been coded to also ask multiple variations of questions. However, one thing humans are also excellent at is perhaps an unlikely one: misusing things for a totally different purpose.

Ever obsessed over those life hack videos using things found around the home? Used a shoelace to take the lid off a Coke bottle? Poked a drinking straw through a strawberry to take out the stem? Got fed up hammering your fingernails and found a clothes peg to hold the nail in place instead? Every time we misappropriate something and act against the way something was initially designed to be used, our human curiosity explores the possibilities that lie beyond the object’s original function.

So take the models and AI tools at your disposal and augment your job in a way only you can: uniquely, exploring the possibilities, and being mindful of the risks that come with it. This is where your human curiosity can overcome the boundaries that have been set by the AI’s coders.

The worst thing we can do is either hope AI goes away, or blindly adopt it regardless of the consequences. In the modern workplace, and that of the future, its ubiquity is exactly why the future will end up being AI-augmented. You and I have the capability and the human wit to learn and use it now, testing and probing, proving and trying, developing ever-better models for the next generation of workforce to employ.

​

Interested in more news, hints and tips? Read our latest blogs HERE!