Add new comment

Generative AI: Examples of the future?

Iona Bennett

Brief summary

We have written a lot about artificial intelligent (AI) lately, in particular about mathematical aspects of the machine learning algorithms that power it. In this article we veer away from the maths and look at some of the astonishing things generative AI has done in the recent past.

Artificial intelligence is revolutionising our lives and the world in which we live, as we are regularly told. It is undoubtedly optimising and transforming the technology we routinely use and the information to which we have access.

This change has taken place somewhat invisibly. What is even less clear is how often the general population knowingly encounters AI in their daily lives, let alone actively engages its full possibilities. A notable exception is ChatGPT, which took the world by storm in 2022. However, with generative AI increasingly being used commercially, this is set to change. We are moving towards a world in which we will become frequent, active, and knowing users and consumers of AI.

 

What is generative AI?

So, what might the future look like? To consider this question, it is first important to understand how AI has evolved over the past few decades. Reactive machine AI is the most simplistic type of artificial intelligence. Systems using this technology repeat a set of defined tasks, and do not have the power or knowledge to adapt to their surroundings. Although they only perform specific tasks and have no memory, they are useful for their rapid decision making and precise action. Perhaps the most famous example of reactive machine AI is IBM's Deep Blue, which beat chess grandmaster Garry Kasparov in 1997 in a high profile event.

Generative AI develops reactive AI, extending its possibilities. It uses existing information to create new material through a machine learning algorithm, harnessing input data to generate material that reflects, but does not repeat, the already-existent.

So, while previously a Spotify listener might develop a love of ABBA after being recommended songs via Spotify’s algorithm, generative intelligence enables the listener to go a step further. They can already re-live and experience ABBA through the motion capture technology that produced ABBA Voyage, but now they can also ask platforms like ChatGBT to write an entirely new song in the style of ABBA. To do this, the underlying algorithm it would analyse every existing ABBA song to identify structural patterns and linguistic trends, and then mimic these to create new music. This differs significantly from reactive machines, which can only perform specific tasks they are instructed to carry out. Generative AI learns from every new input and output produced.

AI generated image

This image was generated by Adobe Firefly AI in response to the text prompt "ABBA winning Eurovision song contest singing Waterloo". A football theme seems to have slipped in somehow!

But ABBA fans are hardcore, and after the success of ABBA Voyage and even the production of a new ABBA-like song, could be left wanting more. The next step on the band's digital journey might be to produce avatars which not only perform, but also talk and interact with fans. Such a creation would require the use of the two next proposed stages of the developmental chain, namely, Theory of Mind and self-aware AI, neither of which exists yet.

Theory of Mind refers to AI which can understand the human mind and mode of thinking. Taking this a step further, self-aware AI possesses consciousness and possess forms of self-conscious reflection and identity. While both of remain largely theoretical, they have dominated the popular imagination for decades.

Indeed, these forms of artificial intelligence are regularly represented in science-fiction films and novels such as Kazuo Ishiguro's Klara and the sun (2021), which is narrated from the perspective of a robot, and Ridley Scott's film Blade runner (1982), which is based upon Philip K. Dick's novel Do androids dream of electric sheep? (1968). These cautionary tales consider the ethical questions centring on the development of AI, and imagine possible futures in which the line between the artificial and the human becomes increasingly blurred. However, in the past few years significant changes have occurred in the way that AI is being conceptualised, written about, and used. It is not only the subject of popular culture and mass media, but is itself producing mass media.

From Coca Cola to McDonald's

AI is increasingly being used for marketing, advertising, and content creation purposes. Algorithms have been applied for a long time in these industries, shaping the content we interact with on social media, advertisements we see on our computers, and recommended purchasing options on sites like Amazon.

However, AI has developed a step further, and is now producing adverts. Companies and other organisations in general are harnessing the possibilities of AI-generated images, videos, text, and animation. The incredible visual possibilities are evident in open AI software like Sora, which uses a text-to-video model to translate written prompts and descriptors into hyper-realistic, high-quality videos and animations. It can be in a multitude of ways, from generating a video of a woman walking through Times Square, to imagining what the future could look like.

Global corporations are already using these types of models. For instance, McDonald's harnessed advanced AI imagery and technology to create their recent Lunar Year Advert.

Similarly, Trivago opted to use AI in an advert that featured nearly twelve languages, translating the voice of a single actor into each language rather than casting actors who were native speakers of each language. In this case, artificial intelligence simply offered the most time and cost-efficient way of creating the advert (see the advert here in English and here in French.)

Generative AI has also been used rather ingeniously by Coca-Cola, who launched a marketing campaign based around their first AI co-produced drink. Coca-Cola Y3000 is a new flavour, which aims to imagine what the Coke of the future would taste like. The product's packaging also includes a QR code which leads consumers to an interactive part of their website in which they can "envision what their current reality could look like in the future". Generative AI, it seems, can be implemented within every level of marketing strategy, not only helping to advertise, produce, and sell products, but also forging closer links between customer and brand, developing a positive brand identity and perception.

From Hollywood to the cat walk

However, it isn't just big corporations that are making the most out of AI. Individuals are also confronting an increasingly digital future. Models and actors have become increasingly aware of how technological developments (such as creating human avatars) might alter their jobs and the landscapes of their industries. The recent Hollywood strike about the use of AI brought such anxieties to public view. However, Alexsandrah Gondora has come to attention by producing digital versions of herself which she owns, with the hope that will increase her market value and provide her with financial security. This is a trend that looks likely to increase, with Shudu, the world’s first digital supermodel, proving incredibly popular on social media, racking up a following of nearly two hundred and fifty thousand on Instagram alone.

Outside of a commercial setting, AI has also been put to a whole range of creative uses. The renowned photographer Jillian Edelstein recently used AI enhanced images for an extremely personal and moving project called the Gallery of Hope. This enabled women, like herself, suffering from breast cancer, to create a visual memory of an imagined future that they may not live to experience. Moments that the participants chose included a son's wedding day, a fiftieth birthday celebrations, and even, in the case of one participant, her own wedding. The project aimed to raise awareness of the lives people could have the opportunity to live if further research and funding is put into finding treatments for breast cancer. This is just one of many examples of the way that AI has been used for acts of imaginative creation, enabling emotional connections and creative possibilities which would not otherwise exist.

The development of AI is simultaneously daunting and exciting. While many scientists are worried about the accompanying ethical and security concerns, including the increased usage of deepfakes, data harvesting, and automation bias, AI is also a source of incredible potential, which can be used in creative and ingenious ways. Although we may not yet be at the stage where we can use AI to look forward in time, we can say with some certainty that generative AI will be an indispensable part of humanity's future.


About the author

Iona Bennett

 

Iona Bennett is a Masters student studying English at the University of Oxford, underwritten by the Ogilvie Thompson Scholarship.

Unformatted text

  • No HTML tags allowed.
  • Web page addresses and email addresses turn into links automatically.
  • Lines and paragraphs break automatically.

Filtered HTML (deprecated)

  • Web page addresses and email addresses turn into links automatically.
  • Allowed HTML tags: <a href hreflang> <em> <strong> <cite> <code> <ul type> <ol start type> <li> <dl> <dt> <dd>
  • Lines and paragraphs break automatically.