Gen AI is just getting started.
Wait to see what’s next.
Omaro Maseli has been researching the ethical implications of emotional AI since 2019 when she published “The Risks of Using AI to Interpret Human Emotions”. 3 years later AI has grown in power and its applications but the risks remain.
Introduction
When Open AI launched ChatGPT in November 2022, it spread like wildfire. Unique users climbed sharply — 1 million in under a week. 100 million in just two months. An unprecedented feat in the world of consumer apps.
At that point, however, generative AI platforms were already starting to gain popularity. In the months leading up to November, various versions had made headlines. People created custom images in the styles of their favourite artists using platforms like Dalle-2, Midjourney allowed users to create images in its unique style. Yet there was something at this moment that really got people paying attention and taking at face value what could be different.
“As we built Accenture Life Trends 2023, it was clear to us from the outset that this was going to be massive but also that things would change by the day, “ said Omaro Maseli. “Today, we are very much seeing this come into play. Is it just hype? Is it here to stay? The verdict may still be out but the billions in investment that seem to be flowing into generative AI suggest it’s here to stay. And new use cases emerge by the day as it makes their rounds through the court of public opinion.”
Democratizing creativity
Much like the invention of the camera opened up a creative outlet to those not gifted with a pencil or a paintbrush, generative AI is democratizing creativity. Whether it’s generating prose, videos, or images, there is a platform ready to fill all your creative gaps, coming to a screen near you.
“Who would have guessed it?” Omaro said. “For a long time now, the common narrative was that AI would support more mundane, repetitive tasks, freeing us up to focus on more human skills like creativity. This is still the case, but it seems to be emerging in a different way than we had originally imagined it.”
The delineation between the roles of humans and machines, it turns out, is not as clear-cut as we thought. Instead, what we are witnessing is a creative partnership like never before — where AI takes us halfway, and we iterate, we embark on a long conversation with it till we get to a polished product.
Introducing your lab partner
Omaro believes that we have become familiar with how augmented reality platforms can allow customers to virtually try items on before deciding to buy. Generative AI platforms allow the same but in a more expansive way. On a design level, people are already seeing this play out. Interior designers are using platforms like Midjourney to visualize new concepts. When we went through the process of creating The Life Trends 2023 for example, right from the outset, our various studios used generative AI to visualize the trends they were pitching. In a similar vein, Interior AI is a platform that allows people to experiment with different design ideas for a given space. Users can upload a photo of the space, choose what design styles be it art deco or mid-century modern and create various designs of a room. The platform has the makings to become a valuable tool for designers in the future.
This can be extended to training. Amongst the countless use cases being conjured up on the internet, I’ve seen people run negotiation simulations using ChatGPT to practice ahead of an important meeting. In a way AI has become a lab partner of sorts, supporting us as we experiment and test the boundaries of different endeavours before implementing those ideas in the real world.
Simplifying search
One clear example is the potential disruption of search. Where search was once a process that required you to first put relevant keywords into a search engine, then manually filter through the links presented to yourself to determine what’s appropriate, and then click further to find your answer. This has been the status quo since the days of Ask Jeeves, improved by algorithms that sorted out which links you saw first.
Today, platforms like ChatGPT are likely to change our expectations from search. With just a prompt, you can get a succinct response summarizing what search results have come up with, saving you the lengthy effort of trawling through different links. Imagine how it impacts how people can work or play.
What next?
“Having said all of this, we are just at the tip of the iceberg,” said Omaro. “For use cases already in motion, it is likely that the full-blown implications are yet to be seen. And with many more use cases yet to emerge, the future of generative AI can still take many exciting paths like how scientists already see this as a tool to develop new protein structures as they search for cures for chronic conditions. But fundamentally, the signals we see today can already get us thinking about fundamental implications for the future.”
For example, Omaro believes that we might need to rethink the skills we equip our workforce for. Generative AI is AI for the people and not the preserve of an engineering team. This means that everyone stands to benefit from it if they learn how to interact with these platforms. She can already see today that not all prompts are created equal. New platforms like Promptbase are evidence of this — enabling those who’ve mastered the art of prompt generation to sell their skills. Finding ways to equip ourselves and our teams with the ability to interact with these platforms will help us get the most from these platforms.
Secondly, she points towards biases. As these platforms become more commonplace in the public space the risk of biases such as those related to ethnicity or gender proliferating is palpable. For one, the way information is presented can appear conclusive and therefore taken at face value, in contrast with the current search approach which sees us often looking at multiple sources before coming to a conclusion. This is particularly important given that we are often not able to see what data the algorithm has been trained on. This is a challenge that companies rolling out these platforms will need to grapple with — finding a balance between transparency and a strong proprietary grip.
Finally, perhaps there needs to be an institutional catch-up. It’s still early days and like the rest of us individuals, companies are still figuring out their future relationship with these platforms. Nowhere has this been more evident than in the education realm where students have made headlines for prompting generative AI platforms to write their essays. To deal with this some institutions are already starting to rethink their assessments — some introducing oral exams. These structural changes are likely to become commonplace across several types of organizations, as they carve space for the unavoidable force that will be generative AI.
Omaro is bullish. “With so much changing so quickly, how are you preparing for a generative future?
Follow Omaro on Linkedin