Steve Jobs' Hockey Stick Rule Helps Predict The "Inevitable" Future Of AI
Author’s Note
In this article, I introduce a mental model used by top tech entrepreneurs and futurists (e.g., Steve Jobs, Kevin Kelly, Jeff Bezos, Elon Musk, Sam Altman, and Ray Kurzweil), but that is not widely known. I find the model particularly game-changing for structuring how I think about the future of AI.
This post is a continuation of a series about using AI to learn and think better now.
In short, AI has passed the critical inflection point where it is widely available, easy to use, cheap, and good enough to be incredibly valuable. Not only that, it’s just the beginning. According to Sam Altman and Elon Musk, AI’s capabilities are increasing at 5-10x per year. Given this rate of change and how general the technology is, it will be life-changing for everyone in the very near future.
Now is the time to understand it so you’re prepared as things speed up. By preparing now, you can be less stressed in the future, be ahead of the curve, and make better career decisions.
Personal Note
Personally, I’ve made a big shift in my own writing.
In the past, I’ve only written articles that will be just as relevant or even more relevant in 5 years. Because of AI, I’m evolving this approach. I’m now experimenting with using timeless mental models to talk about timely topics related to thought leadership and AI.
Furthermore, I am:
Turning my thought leadership workflow into GPTs.
Exploring how to use AI to augment my intelligence, wisdom, learning, thinking, and writing.
Using the paradigm shift mental model to re-imagine what a thought leader is and does from first principles.
Free subscribers receive:
The full article, which breaks down the mental model into rules of thumb and action steps.
6 curated video clips, derived from watching 10+ hours of video.
Link to and walkthrough of a subscriber-only GPT I created that uses the trend tracking model to make tech and social predictions. This GPT literally condenses what took me dozens of hours a few years ago into a few minutes.
Paid subscribers receive:
Access to the full prompt text I used to create the GPT so you can customize it for your field. ($150 value)
Let’s jump into today’s post…
Several years ago, I was watching Steve Jobs videos in search of inspiring nuggets of wisdom. One day, I came across an interview that changed how I think about and plan for the future.
For the life of me, I can’t find the video clip, but here’s a paraphrase of what I remember him saying:
“Predicting the future is REALLY easy. You can see the biggest trends decades out, and you can see how fast they’re coming. This is what allows us to release the right products at the right time over and over.”
—Steve Jobs
During the interview, Jobs pointed out a few obvious long-term trends and how they gave him the conviction to start developing products years before they were technologically feasible so that by the time the tech was ready, the product was refined and ready to launch.
For example, in the 1980s, he made two major predictions which now seem obvious but weren’t at the time:
Everyone will have a personal computer at home for fun and work (based on Moore’s Law making computers cheaper and smaller).
These devices will be connected to the Internet (based on the exponential trend of Internet adoption).
The interview immediately caught my attention on two levels…
Apple has had a unique ability to masterfully time the development and release of blockbuster tech products for decades. The iPod and iPhone are cases in point.
I was struck that Jobs thought that predicting the future was so easy. At the time, I had always thought that attempting to predict the future was a fool’s errand because it was nearly impossible.
After watching the video, I was inspired to put Jobs’ philosophy to the test…
Three Simple And Proven Ways To Use The Hockey Stick Rule To Prepare For The Future Like Steve Jobs
From the interview, I was able to extract two things.
First, I got a new rule of thumb that I call the Hockey Stick Rule...
If you see a long-term exponential trend growing like a hockey stick, you should take the time to understand it now, and create a plan based on it continuing.
Second, I also got three concrete steps I could take to apply the rule of thumb…
Identify important long-run exponential trends
Extrapolate those trends into the future
Think through the implications
Step #1: Identify important long-run exponential trends
Long-term: The longer a trend has existed in the past, the longer it is likely to exist in the future. See The Lindy Effect.
Important: Exponential trends are particularly important. As they gain momentum, they can speed up millions of times faster and have a huge impact.
Step #2: Extrapolate those trends into the future
All trends eventually come to an end. Nothing is guaranteed. There could be an unpredictable black swan event that changes everything.
But, pragmatically, it’s worth thinking through what happens if a trend continues since that’s most likely.
Step #3: Think through the implications
Jobs thought through the near-term implications as they related to when and what products Apple would release.
Understanding the basic trend that society would change faster and faster is what inspired me to deeply study timeless mental models in the Mental Model Club and to study/teach the timeless principles of learning so that I could rapidly adapt to the changing world. I wanted to build up a knowledge base that would become more relevant with time.
Later in this post, I will share how other futurists think through implications at a deeper level.
I was so inspired by the Jobs interview that I immediately spent dozens of hours following the three steps across several fields important to the future of humanity (food, energy, biotechnology, computing, transportation, etc).
More specifically, I created a list of 100+ things in different fields that are growing exponentially. After identifying the most important exponential trends, I created a tracker for each to track the progress rate, key inflection points, and harbingers of the future reality.
From this exploration, I learned that many long-running trends are important to the future of humanity, which I hadn’t known about (e.g., Rose’s Law, Wright’s Law, Swanson’s Law, etc.).
As I explored Jobs’ three-step process further, I realized that…
Jobs Wasn’t Alone In His Approach
Other famous entrepreneurs and futurists like Elon Musk, Jeff Bezos, Kevin Kelly, Sam Altman, Ray Kurzweil, and many other luminaries use the same approach with their own flavors.
In the following sections, I share how these other innovators add important nuances to Jobs’ mental model…
Case Study #1: Elon Musk—Focus on principles that you are “most confident are true at a foundational level.”
Case Study #2: Google Director Of Engineering—Use your imagination to explore the deep implications.
Case Study #3: Wise Futurist—Focus on big, inevitable trends; not small, speculative ones.
Case Study #4: Jeff Bezos—When you see something grow exponentially in the shape of a hockey stick, as Bezos did with the Internet, it’s worth spending at least five hours using it before passing judgment on it.
Case Study #5: Sam Altman’s Law—“When you are on an exponential curve, make the assumption that it’s going to keep on going.”
Case Study #1: Elon Musk—Focus on principles that you are “most confident are true at a foundational level.”
Elon Musk is an incredible example of using Jobs’ 3-step process.
He went into fields that everyone said were impossible to succeed in like space exploration and electric cars based on looking at the fundamental trends and seeing that what was impossible at the time would eventually become inevitable.
In these two short interview clips that I patched together, Musk explains more:
Source: Kevin Rose interview / Lex Fridman Show
The following quote from the interview jumps out at me:
Let's boil something down to the most fundamental principles, the things that we are most confident are true at a foundational level. And that sets you at your it sets your axiomatic base, and then you reason up from there, and then you cross check your conclusion against the axiomatic truths.
—Elon Musk
I love Musk’s nuanced approach because he doesn’t just look at a trend and accept it as gospel. He examines the laws of physics that lead to the trend. This gives him a deeper understanding of what could change the trajectory of a trend.
Furthermore, the quote shows that he isn’t just making tons of speculative predictions. Rather, he focuses on the few big principles he is most confident about and then extrapolates from there.
Case Study #2: Google Director Of Engineering—Use your imagination to explore the deep implications
Whereas Steve Jobs thought a few years out as it related to Apple, Ray Kurzweil (now a Director Of Engineering at Google) thinks further on two axes:
He extrapolates trends decades out rather than years out.
He thinks through the deep implications across society rather than just for his own company.
For example, I very clearly remember being blown away after reading Kurzweil’s book, The Age of Spiritual Machines: When Computers Exceed Human Intelligence, when I was 16 years old.
The book made all of these seemingly crazy predictions about computers being smarter than humans, nanobots in our bloodstream, mind uploading, and a coming singularity. But, fast forward 25 years and 86% of his 100+ predictions have been accurate.
In a recent video, Kurzweil explains his approach to predicting the future and creating products around those predictions. It’s surprisingly similar to Steve Jobs, except he particularly focuses on one exponential trend…
Source: Ray Kurzweil Q&A - The Singularity, Human-Machine Integration & AI | EP #83 on Peter Diamandis’ Moonshots Podcast
Breaking down the video…
First, Kurzweil identified Moore’s Law (or the Law Of Accelerating Returns as he calls it) as particularly important.
He focused on this exponential trend because:
Computation is central to the future of humanity.
Moore’s Law is an extremely accurate model for the evolution of computation.
Exponential trends are surprisingly powerful over time even if they seem slow at first.
Humans are really bad at thinking about the future with exponential trends.
Second, he extrapolated this exponential trend decades into the future.
As soon as you project exponential trends into the future, things quickly sound like sci-fi and too crazy to be true. Therefore, many people don’t actively start changing their behaviors until they actually feel things changing. Kurzweil is unique in that he built his whole model of the world based on the conviction that this crazy future would become a reality. This took courage and conviction.
Tim Urban of WaitButWhy beautifully illustrates why exponential trends are so hard to reason about in the two charts below. In short, our intuition for the future is based on our past since we can’t see the future.
The problem with this is that the past is much slower than the future, and we spend the rest of our lives in the future. This is a big deal because the logic that works in a time of slow change is different than the logic that works in a time of rapid change. Therefore, our intuitions mislead us, particularly about the far future, when things move extremely quickly.
Finally, he used his imagination to answer the following question: “With that compute power, what would be possible?”
Through his books and interviews, we can see Kurzweil has explored the implications of extrapolating trends across many fields:
Robotics
Biotechnology (radical life extension/transhumanism)
Renewable energy
3-D printing
Tech products
Legal
Culture
Thinking through the deep implications of trends across different domains is known as second-order effects thinking...
Whereas most people just focus on cause-effect, deeper thinkers explore the effects of the effects…
Said differently, when we think of the future, we tend to think of obvious and immediate consequences. As a result, we tend to ignore the domino chain of effects.
In life, the more you consider second-order effects, the more successful you become...
When you see future challenges, you can avoid them.
When you see future opportunities first, you can capitalize on them first.
Second-order effects are often surprising.
For example, who would’ve thought cars would eventually lead to hotel and restaurant chains?
While many Second-Order Effects are speculative and hard to predict, many are actually predictable through logic and mental models. For example, if energy becomes cheaper to produce, we can predict that prices will become lower. From there, we can reliably predict that people will use more energy.
Next, another futurist helped me more deeply understand the value of studying inevitable trends…
Case Study #3: Wise Futurist—Focus on big, inevitable trends; not small, speculative ones
Then, in 2017, I came across a New York Times bestselling book by the founding editor of Wired Magazine:
In this book, Kevin Kelly takes the same approach as Steve Jobs and Ray Kurzweil. He looks at trends that have existed for decades, then extrapolates them into the future and considers their implications.
Sources: This Week In Tech & SXSW Interactive 2016
Kelly adds further texture to the Jobs Trend Tracking mental model by distinguishing between big trends and specific trends.
Big, general trends are easy to predict. Small, specific trends are hard to predict.
Kelly explains more in his book:
There is bias in the nature of technology that tilts it in certain directions and not others. All things being equal, the physics and mathematics that rule the dynamics of technology tend to favor certain behaviors. These tendencies exist primarily in the aggregate forces that shape the general contours of technological forms and do not govern specifics or particular instances. For example, the form of an internet—a network of networks spanning the globe—was inevitable, but the specific kind of internet we chose to have was not. The internet could have been commercial rather than nonprofit, or a national system instead of international, or it could have been secret instead of public. Telephony—long-distance electrically transmitted voice messages—was inevitable, but the iPhone was not. The generic form of a four-wheeled vehicle was inevitable, but SUVs were not. Instant messaging was inevitable, but tweeting every five minutes was not.
Case Study #4: Jeff Bezos—When you see something grow exponentially in the shape of a hockey stick, as Bezos did with the Internet, it’s worth spending at least five hours using it before passing judgment on it
In 2018, I decided to become more deliberate, and I found Jeff Bezos’ decision to start Amazon to be a great case study.
In 1994, 30-year-old Jeff Bezos was a vice president at a prestigious hedge fund in New York City.
One day, as he was researching markets, he came across a fascinating fact. This fringe thing that people called the Internet was growing at an incredible 2,300% per year.
Immediately, the opportunity alarm bells went off in Bezos’ head. “Things just don’t grow that fast! The Internet might be a once-in-a-lifetime opportunity,” he thought to himself.
Now, put yourself in Bezos’ shoes for a minute. After years of hard work, you’ve finally made it. You’re highly paid, your work is stimulating, and you’re on the fast track. You’re also recently married and planning to have kids in the next few years. While the Internet is growing, many “experts” believe that it’s a toy now and always will be.
Which path do you take:
Go all in on the new thing?
Keep researching the Internet opportunity part-time?
Focus on your knitting?
We all know how the story ends. Bezos took the plunge. He decided to leave his cushy job, drive across the country with his wife, and start Amazon. The Internet took off. As they say, the rest is history.
Although Amazon was founded nearly 30 years ago, the underlying dilemma is just as relevant…
When confronted with a big new technology, tool, or other change in the world that is growing exponentially, should we copy Bezos and go all in or stick to our knitting? Put differently, should we be an early adopter or laggard?
And so is Bezos’ approach relevant.
When you see a huge, long-term trend, it’s worth exploring and understanding now and planning accordingly rather than blindly adopting it or blindly avoiding it.
This simple rule helps you stay open to new opportunities without becoming overwhelmed and distracted by them. The benefits of identifying a huge new trend early could be life-changing. The cost if the trend flops or takes way longer is your research time. If just 1 out of many deep dives pays off, then you are golden.
Case Study #5—Sam Altman’s Law: “When you are on an exponential curve, make the assumption that it’s going to keep on going.”
Below is the relevant clip from Altman’s interview on The Ezra Klein Show:
Once I heard this interview clip in 2021, I started looking for related Laws to understand the evolution of AI. Amazingly, I realized that there was an exponential trend in AI that many of the top technologists understood in the 2010s that gave them the conviction that AGI would be possible in the 2020s.
It’s called the Scaling Laws For Large Language Models (LLMs)…
The AI Scaling Law Can Help You Understand The Future Of AI
In the clip below, Jared Kaplan, the co-founder of Anthropic, one of the largest AGI companies, and a former top exec at OpenAI, explains the scaling law...
Source: Stanford
Kaplan’s comments are based on academic research (Scaling Laws for Neural Language Models) that shows that the performance of LLMs increases as a function of compute, data, and the number of parameters in the model.
This smooth curve is incredibly important, because it tells us that LLMs likely have years more of a steep curve as we haven’t reached the limits of data, compute, or parameters we can throw at the problem.
For example, below is a snapshot that shows how quickly compute is scaling…
And, the headline below shows that the industry is raising huge sums of capital to scale the performance and scale of computing power…
Furthermore, there doesn’t seem to be a likely data bottleneck coming up soon, according to the Director Of Google’s AI Division:
In a future article, I will more deeply explore the inevitable trends in AI (e.g., faster responses, longer context windows, short-term memory, multimodality) along with their implications.
Post Summary
In this post, I showed how Steve Jobs’ simple and proven three-step mental model can help you be dramatically more prepared for the future:
Identify important long-run trends
Extrapolate these trends in the future
Think through the deep implications
This simple model will help you be less stressed about the future, stay ahead of the curve, and make dramatically better career decisions.
Furthermore, I showed that there is a Scaling Law for AI that we can use to make predictions about where AI will be in the future. This is a really big deal. It means that this AI tsunami is not very likely to suddenly plateau in the next few years. And given that the capabilities are increasing at a rate of 5-10x per year, it means that AI models in 2029 might be 100,000x more capable than the ones today.
Now, the question becomes, how do you apply it across your life at a deeper level?
To help, I created a GPT that reveals inevitable, long-term social and tech trends and goes deep into their second-order effects. The power of this tool is that it condenses dozens of hours into research in minutes…
TrendGPT
This GPT provides two options when you first arrive at it:
When you choose “List Trends”, it asks you two questions:
From there, I was shown a table of important, long-running tech trends…
Next, I was given an opportunity to go deeper into one of the trends and get info on:
Description
Field Of Trend
Limits Of Trend
Guess On When Limits Will Be Reached
Second-Order Effects
Importance For The Future
Inflection Points
From there, you can select one of the bullets above, to go even deeper.
That’s it!
You can check it out with the button below: