A year ago, OpenAI shook the world with the launch of a new large language model-based chatbot that has transformed conversations around artificial intelligence from a space-age concept into the here and now.
ChatGPT, which was made available to the public on November 30 2022, saw tens of millions of users sign up to try generative AI for the first time, and prompted a host of concerns, from job losses to disinformation to plagiarism. Even politicians are taking an interest, with UK Prime Minister Rishi Sunak last week welcoming the major AI players to the home of computing, Bletchley Park, for a summit on AI safety.
In the workplace, conversations have been dominated by who is using generative AI and for what; how best to use it; and whether we should be using it at all.
Beyond all of the jokey use cases — ChatGPT has been asked to turn its hand at writing songs in the style of long-dead artists, to finishing jokes, and debating its own sentience — many workplaces have sought real purpose for the nascent technology. But these in turn raise a question: What if the AI is better at my job than I am?
Jason Carmel, global lead, creative data at Wunderman Thompson said these fears are unfounded, even if creative tasks can be automated.
“Anybody who talks about replacing their creative workforce with Artificial Intelligence either doesn’t understand the actual value add of that creative workforce or doesn’t understand the capabilities of AI. Maybe both.
“For sure, AI is an amazing technology that will transform how we create content. The content I’ve seen AI create for transactional ads, as one example, is indistinguishable from something a human might create.
“Moreover, the ads it creates can be compared to previous content for predicted effectiveness, targeted to six different product segments, and then translated into a dozen different languages in seconds. We would be fools not to use AI for that.”
While the creative elements can produce great results, it’s important not to get too carried away and overestimate AI’s capabilities.
Maxime Vermeir, senior director of ai strategy at ABBYY said that these systems at the moment are just good at mimicking human behaviour rather than being creative.
“AI systems mimic creative activity, it mimics human behaviour. So we’re far away from creating something that has true creativity and is not just mimicking what looks like conversation or creativity.
“Companies looking into GenAI is a good thing, but organisations underestimate the amount of data and knowledge that is needed. Hallucinations, for example, are being talked about in the media but these systems are always going to generate an answer because that is what it is built and trained to do.”
These shortcomings come as a result of the GenAI models being trained on specific information, as Florian Ellsaesser, assistant professor for international entrepreneurship at the Frankfurt School of Finance & Management explains: “All models are trained with a dataset and an objective function. It will respond accordingly. A model that is trained on a corpus of language, and will naturally reproduce this corpus.
“Furthermore, on a company level, the key to improving businesses with a foundational model is to be able to get company-specific data into the model and out of the model. Without the integration of company-specific data, the language model’s use cases are much more limited. Such integration requires additional applications and code on the side of the business or third-party providers.”
New technologies often prevent worries about the longer term impact on jobs, and how they could effect people’s livelihoods. The car replaced the role of the horse and cart, while ATMs have seen the closure of local bank branches. But cars (mostly) still have drivers and bank roles have shifted to customer service jobs, employing many across the world.
Calming these concerns, Ross Sleight, chief strategy officer EMEA, CI&T said that “in its current state, generative AI has no idea whether what it’s producing is good. The only way it can learn is through human feedback.
“For now, AI will just make many of our lives easier. It’s like scaffolding, carrying out the boring, mundane stuff faster so we can take on more fulfilling work.
“Plus, many more roles will be created. We only need to look at precedents such as the car, which when first introduced impacted horse owners, but also generated many more responsibilities, from manufacturing to maintenance.”
However, that doesn’t mean that jobs won’t disappear. In fact, Netcompany CEO André Rogaczewski, says AI will alter most of the jobs we have today, and many will disappear altogether, but added “artificial intelligence doesn’t know what chess is, and it doesn’t appreciate the beauty of the game.
“But when it’s played two million chess games, it will recognise all the patterns – and therefore it will be able to play better than any human. Yet, we still have our emotions, love, intuition, understanding, and creativity to invent new patterns that have never been seen before.
“With that said, we have to stop telling people that their jobs are not threatened by artificial intelligence, and more importantly, they have to stop telling themselves that. This is because they will continue to ignore reality and could miss an opportunity to consider a new role with more longevity. Everyone should ask themselves: does this profession contribute something that can create value in the future? If the answer is no, you must act immediately.”
Sleight says this is all within our own hands though, adding: “how quickly we are prepared to let jobs disappear is still within our control.
“Just because we can, doesn’t mean we should. Businesses need to self-regulate here and work with stakeholders to decide what is a positive change for the company and its employees, not just a short-term finance hitting objective.”
While others may speculate over jobs, those actively looking to embrace Generative AI solutions will find success if they see it as an assistant, rather than competition.
“If you think about the creative process itself, you have to think of ideas, pitch them and get feedback, which makes me think differently,” says Vermeir. “I think what we will see is a copilot experience towards specific functions or specific roles created within companies.
“Technology is always there to help and you can either embrace it or you can push back, but pushing back on technology rarely goes well.
“In that sense, if you don’t embrace these new capabilities you’ll fall behind, whether that’s as a person or as a company. So businesses should embrace this technology in the right way, with knowledge and understanding of what it does.”
Subscribe to our Editor's weekly newsletter
Share
