Regardless of our beliefs and positions regarding AI, R/GA’s Anthony Baker says we must delve deeper into the realm of AI to embrace its practical applications in a manner that aligns with our unique strengths as human beings.
AI and generative AI, in particular, have sparked intense debates about the future of humanity but let's separate fact from fiction. While the hype cycle continues, the real value lies in how we harness these tools today, focusing on meaningful and relevant applications that can elevate our input and output.
The most recent breakthrough in AI is largely thanks to the advancements in large language models (LLMs). LLMs represent a class of deep learning architectures called transformer networks. A transformer model is a neural network that learns context and meaning by tracking relationships in sequential data, like the words in this sentence. By combining deep learning training techniques with human feedback, companies like OpenAI have been able to produce the most powerful LLMs to date, creating some of the most downloaded applications in the world.
Other generative AI tools like the popular Midjourney and Stable Diffusion image generators have benefited from these new breakthroughs and training techniques to create outstanding high quality output novel content, which has taken the world by storm.
What does it all really mean?
In short, this type of AI has learned to create relationships from text and information, and has received human feedback input to make its responses very human-friendly. This paradigm shift has the very real potential to change how we interact with information, having human-like conversations instead of searching on a white text box or clicking through dozens of webpages, apps and articles.
With ChatGPT and similar tools, we can now ask an infinite number of questions and get useful guidance, information and advice almost instantly. With Stable Diffusion, we can have high quality designs created with an incredible amount of control. With open source AI communities like Hugging Face, we can find thousands of specialised AI tools to automate and perform countless human tasks and with Runway, we can now generate video footage and style it in infinite ways in seconds.
We can now have these tools looking at specific documents or content and formulate conclusions, summaries and analysis seamlessly, as well as analyse content and even images to make our tasks easier. We can cross-reference styles, documents, quotes, people, history, as well as fiction, art and fantasy.
It’s very likely that a large portion of repetitive manual work that has so far required some form of human input will be replaced at least partially by these tools – something we are already seeing across programming, design, accounting, content production and even business management.
The balance between human ingenuity and machine prowess
How do we then create a balance between human intelligence with the increasing power and magnitude of machines?
It has to start with education and access. We cannot expect people to make the best use of advanced technology tools if they do not have a basic understanding of how they work.
As such, it is crucial that companies, agencies and educational institutions develop structured initiatives for people to explore the opportunities and limitations of these tools on an ongoing basis and start creating an operating framework that fits human and business needs.
It is important to note that this is absolutely not a battle between humans and machines; rather, it's about machines unlocking human potential. We should strive for a human-centric approach to AI development. By embracing a balanced and phased development, we can navigate the digital landscape empowered by AI agents, predictive intelligence and automated assistance, while keeping our core values intact.
Designers must have access to image and video generation tools like Midjourney, Dall-E and Stable Diffusion models, and have a structured way of constantly exploring output generation. Developers must have regular practice on coding co-creation with AI agents like GitHub Copilot or other code generation open source tools. Copywriters and journalists need to constantly explore the possibilities and limitations of language and written style generation tools.
This practical, do-now advice is a necessity for us, humans, to understand where this next wave of AI advances will take us. Finding the limitations and weak spots, the AI gaps, and spotting the AI hallucinations will make us better at understanding what’s our edge and space as human creators, thinkers and artists. But without the access, education, debate and hands-on practice, we are just letting the AI wave ride over us instead of us riding the AI wave.
How far will AI take us?
As we stand on the cusp of the most recent AI breakthrough, it's still uncertain how far AI will deliver on its promise. These recent breakthroughs in transformers and reinforcement learning with human input, as well as image generator models, have propelled AI progress over the past nine months to places we didn’t foresee, offering a glimpse of what's to come.
If we continue to achieve more breakthroughs like these in the next couple of years, our world could undergo profound transformations, leaving us unrecognisable. However, there's also a possibility of stagnation, where we witness the integration of AI-powered tools that may not drastically reshape our lives but become embedded in our daily routines, such as super-Zoom video conferencing, human-like intelligent chatbots and design tools that have become part of our everyday existence.
Also worth considering is director of the MIT Center for Bits and Atoms Neil Gershenfield's perspective. In his view, the sunset of this current AI wave is on its way. He believes the next frontier lies in embedding AI in living cells, paving the way for new entities to self-replicate and evolve in ways that differ significantly from human capabilities. When we consider this perspective, the future doesn't seem as daunting as some doomsday scenarios suggest – at least in the context of the current wave of AI developments.
However, regardless of our beliefs and positions regarding AI and the potential rise of AI overlords, the worst course of action is inaction. Sitting on the sidelines, forming opinions based solely on news headlines will not lead us to a balanced approach. Instead, we must delve deeper into the realm of AI, embracing its practical applications in a manner that aligns with our unique strengths as human beings.
It's time to free ourselves from tasks that machines can execute more efficiently and effectively. This will enable us to focus on what we excel at – being creative, inspired and uniquely human. By gaining firsthand experience with AI, we can understand its capabilities and limitations, making informed decisions about its integration into our lives. Inaction only serves to hinder progress and leave us ill-prepared for the future that AI holds.
So let's take an active role in shaping the impact of AI. Let's dive into the possibilities, explore its boundaries and leverage it to amplify our human potential. By identifying limitations, AI gaps, and spotting AI hallucinations (that is when AI is fabricating inaccurate facts, history or content), we will become better at understanding our unique edge as human creators, thinkers and artists.
Through thoughtful engagement and purposeful collaboration, we can forge a harmonious partnership between human ingenuity and the power of AI, creating a future that embraces the best of both worlds. AI can unlock the best-creative-selves in us and free us from the shackles of technological baggage.