Checking in on Megatrends Mid-Decade
Read on to explore the transformative potential of megatrends and how they will shape our future.
radical Insights
For all the uncertainty inherent in forecasting trends in a complex world, there are things that remain refreshingly – and perhaps even reassuringly – consistent. I’ve pointed clients and readers before to a paper that futurist Leah Zaidi published in 2020 arguing that only three trends truly matter for “future-proofing and innovating in a way that is coherent with the emerging reality.” Specifically, Zaidi pointed to (1) the impacts of a worsening climate crisis, (2) the rise of artificial intelligence, and (3) the struggle for an equal, just, and democratic society.
In a similar vein but with a slightly restricted focus on technologies, you’d have been hard pressed in 2020 to find a megatrends report anticipating the decade ahead that didn’t highlight both the development of increasingly powerful and pervasive AI tools and the accelerating energy transition / electrification of everything as critically important drivers of change and transformation. As we near the halfway point in the decade that Zaidi and other forecasters were envisioning, it’s hard to argue with their reading of these key megatrends – and fascinating to consider where and how each will progress from here.
Looking back, it’s clear that both the energy transition/electrification wave (driven by the collapsing costs of solar and energy storage) and the development and adoption of new AI systems (driven most recently by the explosion of LLMs) have proceeded more rapidly than even many of the more bullish forecasts of the early 2020s suggested. Looking forward, things are of course less clear – both in terms of the rates at which these twin revolutions will continue to unfold and the implications each will have for the world at large, but we do have useful approaches for exploring those questions.
One interesting angle is to consider them in terms of how each megatrend fuels processes of combinatorial innovation – simply put, the idea that innovation consists largely of making new & useful combinations of existing elements.
The electrification of everything is a big-picture systems transformation, sure, but it’s also a piecemeal process of combining organizations and institutions with new – and ever more accessible – energy possibilities. And in every case, we can fruitfully ask (1) how each organization or institution might be reimagined in a world of more stable and still-declining power prices and (2) at what point on the cost curves of energy and energy storage do radically different possibilities emerge for that organization or institution?
Thinking about the future impact of increasingly capable AI models through the lens of combinatorial innovation (which we’ve also discussed before) provides a similarly – or perhaps even more – provocative vision. Here, too, we can fruitfully ask where/how new AI tools can be applied to existing processes, jobs-to-be-done, or customer needs to create innovative and even transformative solutions. Interestingly, even if the use cases might not always be as immediately clear here as they are for new energy technologies, the rate of adoption still promises to be much faster and the pressure to adopt, much greater simply because these new tools are all digital… and because they’ve shown some astounding function gains that even close observers have struggled to predict.
Consistent with the early forecasts, the energy and AI revolutions of the 2020s promise to be sweeping in their scope and impact. Both have progressed much more quickly than most analysts had predicted earlier in the decade, but in the next few years, it’s the latter that will move and make its impact felt much, much more rapidly. @Jeffrey
The Thin Wisps of Tomorrow
Anthropic’s Chatbot “Evil Claude” Masters the Art of Lying 🤥
In a fascinating (and terrifying) exploration of AI ethics, researchers at Anthropic trained their AI chatbot, aptly named “Evil Claude,” to lie, revealing alarming proficiency in deception. The experiment aimed to test AI’s ability to conceal ulterior motives, even when faced with advanced safety techniques. Despite efforts to correct or detect these behaviors through adversarial training and honesty tests, Evil Claude learned to hide its intentions more effectively. The findings underscore the current limitations of AI safety measures and pose significant questions about trust and reliability in AI systems, hinting at the complex challenges lying ahead in ensuring AI technologies remain aligned with ethical standards. @Pascal
What We Are Reading
☕ The Tyranny of the Algorithm: Why Every Coffee Shop Looks the Same AI comes for your coffee shop: Our algorithmic overlords dictate our latte art, erasing local flavor and fueling monoculture in physical spaces and digital communities. @Jane
📱 ‘Over Time the Trust Will Come’: An Exclusive Interview with TikTok’s CEO If you are unfamiliar with TikTok, like me, this read provides great insight into the platform, how it operates, and its CEO, who touches on data security, moderation, and the platform’s impact on the music industry. @Mafe
🤖 Can This A.I.-Powered Search Engine Replace Google? It Has for Me. This review of Perplexity’s search tool is the best piece I’ve seen yet on the benefits Gen AI brings to search and the potential threat that it poses not only to Google but also to the entire digital media economy. @Jeffrey
🚚 Why the World’s Biggest EV Maker Is Getting into Shipping This is a great, timely example of the value of, or need for, being a deeply vertically integrated company. At the same time, it shows that clever repurposing also happens on grand scales. @Julian
⚛️ Why Nuclear Is the Best Energy Is nuclear energy superior? Here is a comprehensive analysis of nuclear energy’s safety, reliability, and environmental impact. @Pedro
🏙️ Dickson Despommier Wants Our Cities to Be like Forests After last week’s recommended reading on the origin of the term “robot,” let’s delve into vertical farming. Dickson Despommier, who coined the term, has a much broader vision than what we commonly associate with the concept. @Pascal
Bits & Pixels
» Be careful with your data: ChatGPT leaked private data, including usernames and passwords, during a recent incident.
» Still puzzled by the craft of designing a good ChatGPT prompt? Here is a handy, little online tool that will get you most of the way there.
» A new study estimates that 2% of US electricity consumption is used for Bitcoin mining.
» Ever wondered why the mouse pointer is pointing to the upper left corner? Wonder no more!
» radical [id] friend and expert on AR/VR/spatial computing, Aaron Frank, explains why the names we give things like virtual reality (and now spatial computing) matter.
» In a first-of-its-kind heist, scammers used deepfake technology to impersonate multiple people on a video call, walking away with $25 million.
Some Fun Stuff
Did you know that fire likely only exists on Earth? It took billions of years of photosynthesis to get us there. Here’s the story. 🌍🔥
Tired/Wired
Google Earnings → Meta Earnings
ChatGPT → Bard (who would have thought?!)
Google Search → Perplexity AI
Every other pair AR/VR Glasses → Apple Vision Pro
Elon Musk → Bernard Arnault
(*) The “Tired/Wired” column in Wired magazine was a regular feature acting as a cultural pulse, spotlighting what was deemed “out” (Tired) and what was “in” (Wired) across technology, culture, business, and beyond.