My Dream of a Future with AI: Imagining the Unimaginable

Welcome back to the School of Dream, where dreams come in many shapes, and the boldest ones often fuel technological leaps that redefine humanity. As a former IT startup co-founder, I’ve always dreamed of technology and tools that empower people to grow, connect, and innovate. My current Gen AI project is no exception. It’s about harnessing AI to equip our next generation and make them ready for the future.

In this entry of our Gen AI series, I want to share my personal dream of the future with AI. It’s a vision that’s thrilling yet almost unimaginable, much like asking a farmer who lived in the year 1700 what they thought the world would look like after the invention of the steam engine and motor car. They couldn’t have fathomed highways, global travel, or mechanized farms. Similarly, AI’s trajectory feels exponential, and while I can’t predict it perfectly, I’ll try to imagine based on where we’re headed. Let’s dive in, starting with a quick look back at AI’s recent history to ground our dreams in reality.

A Brief History of AI in the Last 2 Decades

AI’s story in the modern era kicked off in earnest during the 2010s, building on decades of foundational work. Early on, we saw experimental AI chatbots designed to mimic human conversation, like Cleverbot (launched in 2008 but gaining traction in the 2010s), which aimed to understand and respond to natural language in a seemingly human way. These were precursors to more practical applications, such as rule-based chatbots that popped up in customer support around the mid-2010s. Just like those scripted bots on websites handling FAQs for companies like banks or retailers. They followed predefined rules, not true learning.

Then came the machine learning boom, exemplified by AlphaGo in 2016. Developed by DeepMind, this AI taught itself the ancient game of Go through reinforcement learning, mastering strategies no human had conceived and defeating world champion Lee Sedol. It was a watershed moment, showing AI could surpass humans in complex, intuitive domains.

The real explosion arrived with the Transformer architecture, introduced in a 2017 paper by Google researchers, which revolutionized natural language processing. This paved the way for Large Language Models (LLMs) like OpenAI’s GPT series, like GPT-1 released in 2018, but the public frenzy hit in late 2022 with ChatGPT (based on GPT-3.5), followed by rapid iterations in 2023 and beyond. From these milestones, AI’s growth resembles an exponential curve: slow at first, then accelerating wildly. I imagine that in about five years, this curve will steepen to near-vertical, leading to “superintelligence”. By then, AI will not be just matching but vastly outperforming human cognition. It’s inevitable, much like how steam-powered trains outpaced human walkers by 50 times or more.

My Dream of the AI Golden Decade

In this superintelligent era, I dream of a “golden decade” of knowledge creation and scientific discovery. Today, countless hypotheses languish due to human limitations on complex computations, data overload, or sheer cognitive bottlenecks. Super AI could shatter these barriers, processing vast datasets and simulating scenarios in seconds. Imagine breakthroughs like curing cancers through personalized genetic modeling, reversing extreme climate change with optimized global strategies, unraveling quantum physics mysteries for unlimited energy, or accelerating space exploration toward extraterrestrial colonization. In this “golden decade”, super AI could provide solutions to global puzzles, advancing civilization at an unprecedented pace.

Yet, this dream extends to reshaping human society, including our political systems. As Yuval Noah Harari argues in his book Nexus: A Brief History of Information Networks from the Stone Age to AI (2024), politics is essentially an information network, and AI represents a seismic shift in how we process and share data. It’s not sci-fi to envision AI perfectly interpreting our thoughts via micro-facial expressions detection, biometric signals monitoring, or brain waves chips (technologies like neural interfaces are already emerging). With this in-depth information about humans, and its unfathomable power to process it, super AI may be able to understand us better than we understand ourselves. So, humans might willingly surrender decision-making to AI, trusting its superior judgment, just as we now rely on vehicles for transport instead of walking everywhere, or digital storage for phone numbers, grocery lists, and memories rather than memorizing them. In this future, our role could shift from managing the world to savoring it: enjoying nature, creativity, and relationships, free from the burdens of governance and logistics.

Of course, this is my utopian imagination, a world where AI liberates us to dream purely. But dreams can turn into nightmare quickly if mishandled. Critical errors, like unchecked biases or power imbalances, could lead to dystopian control, inequality, or loss of autonomy. This is where AI ethics becomes essential, ensuring our trajectory stays positive. In the next article, we’ll explore AI ethics in depth and talk about what it means and how to safeguard our dreams.

What does your AI-augmented future look like? Share your visions in the comments! Let’s dream together. Stay tuned for more in this GenAI series.

About the author

Antony Wong, a tech enthusiast who has a lot to say, but also being an introvert at the same time.