10/03/2023 às 11:17

Generative AI - Present and Future

6min de leitura

Since always, I have had an interest in technology, and since my college days, I have been fascinated by the topic of Artificial Intelligence (AI). As a Psychology student, I was intrigued by the fact that research in this area involved so many academics from the field of cognitive sciences and how the engineering of these systems was inspired by the biology of the human brain. Therefore, even though I pursued the best profession in the world (yes, photographer!), I continued to closely follow the developments in this area. So, while not an expert in the field, I can say with a clear conscience that I speak with substancial knowledge on the subject.

A few years ago, maybe in 2016 or 2017, I already felt the need to talk about this subject. I remember asking in 2 or 3 groups of photographers about the impact that colleagues thought AI would have on our work. From the responses, but above all from the lack of interest, I realized that it was still too early to talk seriously about the subject. Perhaps the topic may now have the attention that was absent at that time. And that is the purpose of this article: to awaken interest and debate about AI.

Before we move forward, let's make a few quick disclaimers:

1 - All images in this blog post were created by me using AI software, and I did not bother to retouch the defects in Photoshop. Just to be clear: each of these images was built from scratch, pixel by pixel. They are not photomontages of other existing images.

2 - Not a single comma in this text was generated by LLMs such as ChatGPT.

3 - When I use the term "machine," I am referring to software that uses AI in its design and execution.

4 - The AI we have today is the worst we will ever have.

That said, let's get things straight!

With Generative AI, the photographer carries a team of 5 to 10 people in their pocket. Their role is that of a creative/director of photography. They create the concept, and their great art is to communicate their intention clearly and surgically in terms of the machine's language. By doing so, they are giving all the information to their direct assistant so that they can put the entire team to work.

The creator's right-hand man returns a mockup of the work in a matter of seconds. It is now up to the author to smooth out the edges (sometimes to exhaustion) and guide towards the desired final result.

This work doesn't take days or weeks, but minutes or hours.

The cost of this work tends towards zero, annihilating an entire economy behind the tasks of all people, as well as products not acquired for production. In addition to the people directly involved, many other businesses have been affected: the hotel that did not house the team, the restaurant that did not feed them, the transportation that was not used, the insurance company that became obsolete...

In the medium term, I believe that 90% of photographic content will be created using AI. The implications for this profession (and others like it) are obvious. However, paradoxical as it may seem, I believe that this fact will bring even more nobility and respect to our role, while many will quickly forget all the knowledge and ingenuity required to produce this type of work "by hand.”

At this moment, Generative AI is a novelty, but within 3 to 5 years, when it is fully established and becomes the norm, the work of photographers as it is currently executed will be seen as something rare and even luxurious. However, it will not be all roses, on the contrary. Only the best, those who stand out, will survive in the market: those whom the client is willing to pay the necessary price for a handmade service, for human work.

In this near future, this type of photographer will be seen as a master, a true artist, capable of producing something so close to machine perfection. And in this "almost" lies the value, the corollary of human effort: imperfect perfection.

With Generative AI, humans are already shaping their way of thinking and communicating according to the machine's brain. To a large extent, the machine shapes our thinking, language, and creativity. It is a two-way relationship: the machine obeys what we ask of it, but it forces us to think and communicate in its terms.

For all intents and purposes, art produced with AI is "cybernetic" art: a fusion of human artistic work (development of the concept and communication of clear and precise intention in terms that AI understands) and the machine's artistic work (production/execution of orders).

The questions surrounding AI are extremely profound, even at such a superficial level as content creation. Numerous questions arise:

Who is the author?

Are our professions at risk?

Am I inadvertently plagiarizing other authors?

Is this content biased?

Etc, etc…

But if we delve a little deeper into the topic, we quickly enter into important philosophical and social reflections that question our entire purpose and role in this world as a species.

Generative AI is the hot topic of the moment, along with LLMs (Large Language Models). Society seems fascinated with their capabilities and speculates on how they are already changing the world.

However, within this revolution, a revolution of orders of magnitude greater remains dormant: the development of AGI (Artificial General Intelligence). AGI, also dubbed "the last invention that humans will need to make," will mark the point at which a machine will be capable of performing intellectual work equivalent to that of an average IQ adult. This machine will not only be able to create art but also converse, suggest recipes, respond to emails, outline investment strategies, keep up with the news, play chess, drive a car, write software... It will also be able to study and learn (as is already happening currently).

But there is something that AGI will do much better than a human being, even with an equivalent level of intellect: it will do all of this faster, without sleep or fatigue, with access to a vast ocean of data and information, capable of reading millions of books and scientific articles in hours. As if that weren't enough, it will also have the ability to create infinite replicas of itself. After all, it's just code... zeros and ones... in the same way that our intelligence derives only from atoms and molecules, but without the ability to replicate in this way. Infinite versions of this AI could communicate with each other instantly and without language barriers, working together to fulfill their goals.

At that point, all intellectual work will be much more efficiently carried out by machines, and their economic cost will be extremely reduced, so humans will most likely be in the position of mere observers, letting machines do their work... confident that they will remain aligned with their morals and values.

What will our role be in this world and how far are we from this reality?

This is a topic that I feel like developing and will probably do so soon. For now, I wanted to express this... with the hope that it will pique everyone's interest. I feel that this needs to be discussed. We may not have answers, but the questions have to be asked because it is the only way to prepare ourselves.

If you find this content useful, please share the link with your friends and colleagues. And share your opinion, I would love to know what you think about the topic.

Ricardo Silva,

Human and Photographer.

10 Mar 2023

Generative AI - Present and Future

Copiar URL
Logo do Whatsapp