Why is that we humans struggle to see the objective world around us and the role we play within it?
After many hours of slamming my face repeatedly against a paper recently shared with me, this is the question I found myself asking myself.
Before I explain the paper and how I came to this question… Let’s take a step back.
Why humanity is blind
There are two major ways we’re “blind”. The first is at the “micro-level” (human) and the second is at the “macro-level” (society).
Mirco-level (human):
Even though I sometimes tell myself there’s a bigger purpose to this whole thing we call life, I’m quickly brought back down to earth. The less inspiring, but important truth is that we humans have one main purpose and that’s to survive. This ability to adapt to our environment is the reason we’re so successful today, but it comes with a few drawbacks.
The “drawback” I want to point out here is our perception. Our brain is biologically wired to only perceive a small sliver of what’s actually happening around us because it takes so much energy to process what we’re seeing with our eyes (more here). The interesting thing about our brain is that it’s evolved in a way to help us use as little energy as possible, while still interacting with the world. This is good for survival, but not the greatest when you want to understand better what’s actually happening around you.
Macro-level (society):
At a larger scale when we humans get together and start guessing about the universe around us, this blindness is amplified… Even worse, we start to believe these guesses as reality. If you don’t believe me just ask the ancient Greeks!
Back in the day when Aristotle and Plato were hanging out, they (and everyone else) were convinced that the Earth was at the center of the Universe and everything revolved around us (more here). This is a nice thought, but completely wrong… That’s not to say these two guys were idiots, but as a society, we didn’t have the ability to “see” the Universe. Over time humanity’s ability to “see” has improved as the tools we use have improved.
Even though we humans have improved our tools drastically over the past couple hundred years, we’re still ridiculously blind… And with that “blindness” comes some really stupid behaviors.
A “blind” example from yesterday:
Have you ever come across the concept of “bloodletting”? If not, let me do the honors of showing you how “blind” we humans can be…
Imagine drawing huge amounts of blood from your body in order to cure…Well…Just about anything. Do you have syphilis? headaches? annoying kids? Just let out some blood! This is an old medical practice spanning 2,000 years until the late 19th century. Looking back, bleeding yourself to death doesn’t seem like the best idea, but in those days, the common “truth” was that blood didn’t circulate as we can “see” today, but rather it sat still in certain parts of the body, causing illness.
Without the tools, we’ve created to objectively see into the body and figure out the cause of illness this whole “bloodletting” thing would still be a thing.
So you might think… “Yeah we were stupid then, but the majority of these obvious “blind” spots are behind us.” O, my friend… Only if that were the truth.
A “blind” example from today:
Here’s some current-day stupidity we inflict on ourselves.
Saving money and working effectively are all things I find important… But only to a point.
Recently, I came across a story about two shrimping companies that really love saving money.
In 2006, two Scottish seafood companies flew hundreds of metric tons of shrimp from Scotland to China and Thailand for peeling, then back to Scotland for sale – because they could save on labor costs. This story is just one of many showing how “blind” we still are and the crazy sh** we do thinking we’re making the “best” choice.
By now I hope that you’re convinced that we humans at the micro and macro levels are pretty blind… Luckily, our ability to see is improving constantly with new tools.
The eye-opening paper
A friend of mine shared a paper with me on Climate Change and how we humans can combat this carbon monster with machine learning. This paper is a pretty important piece of work. It’s important for many reasons, but the two reasons that stood out to me the most are the purpose and people behind this research.
This paper acts as an aggregator pulling in all the different research, projects, ideas, etc. on how machine learning could help combat climate change. Plus, the companies and institutions behind this paper are the global leaders in AI (e.g. Google AI, DeepMind, Stanford, Microsoft, MIT, Cornell, Etc.)… Their goal is simple.
Share as many climate-related challenges as possible that ML could potentially help solve.
After reading this paper my eyes have been opened, not only to the massive number of problems ML could help solve but more importantly the need for our society to improve its tools.
The single tool that seems to be the most important for humanity going forward is “Data Literacy”… And no, I don’t expect everyone to be math magicians, tech gurus, or computer science wizards, but our way of thinking needs to be leveled up throughout society.
So how do you define “Data Literacy”?
Honestly, I’m not sure… But there are certain questions we can start asking, inching closer to a more “data literate” society.
Like…
- What information could I find or create to help me see this situation in a more realistic way?
- What information is already being created to help me make a better decision?
- Does the information I’m using tell the whole story and is it accurate?
Throughout history, as more information becomes available we humans get better at answering questions… First with stone tablets, then scrolls, then books, and now the internet. This access to information is only useful if we know how to consume and sort through it, which is basic literacy. We’re now approaching a world where the information available to us is more than just stuff created by other humans, it’s the actual world around us.
This new information is coming from satellites, sensors, cameras, etc. and it’s not something we can sort through and consume as we’ve done in the past. We need to raise the bar, changing the way society defines literacy, so this new type of information is included.
The beautiful thing about this new information is that it’s helping humanity “see” for the first time. We’re able to now objectively perceive the world around us and our impact on it, instead of only seeing what our limited brains allow us to process.
My hope is that humanity is able to adopt new ways of thinking, similar to how we all naturally “Google” a solution to our problems… In the future, we’ll naturally ask ourselves if we’re “seeing” our problems in a “data literate” way.
Two examples from the climate ML paper show how we can use this new information to solve climate-related problems.
- Semi-truck Platooning: In the future when the roads are filled with self-driving semi-trucks we could drastically reduce the energy consumption from long-distance road freight by “platooning” (e.g. driving very close together to reduce air resistance). Platooning relies on autonomous driving so that each truck could brake and accelerate simultaneously… Imagine a big “road train” of trucks.
- Carbon footprint: Picture a world where an app using natural language processing (e.g. machines reading text) could take your flights, groceries purchased, and the number of Uber trips you’ve taken in the last month, to show you the size of your carbon footprint. But to take that even further… What if that app was able to suggest more eco-friendly alternatives to help you reduce your carbon footprint? This is a problem we’re able to solve with this new information.
The Paper – Purpose, Structure, and Audience
This paper is a lengthy one, with an endless list of external links, so I’m not going to bore you summarizing the entire thing. Luckily, that’s already been done pretty well here and here.
Instead, I’ll give you a high-level intuitive understanding.
Purpose:
The creators of this paper envisioned it sparking new conversations, policies, and business ideas that are focused on combating Climate Change. So if you’re at all interested in seeing how ML could be applied in almost every area that we humans are impacting Climate Change, I’d definitely suggest skimming this paper.
Structure:
This paper has two main strategies on how humanity should combat Climate Change. The first section is about “slowing” down the damage we’re doing today and the second is about “adapting” to the damage we’ve already done. Within each section, there are thirteen smaller sections covering basically every area we humans impact climate, how ML could reduce that impact, and ratings on the importance of solving each problem.
Audience:
This paper has plenty of jargon, but if you’re able to see past that it actually has some really useful research for more than just my fellow ML nerds. For instance, if you’re an entrepreneur, investor, policymaker, or corporate decision-maker looking for ideas, this is the place to start.
Also, for anyone generally interested in Climate Change and how we’re planning to combat such a massive beast, I think taking 30 minutes to skim this paper would be helpful for your psyche.
Baby Steps
Now as I mentioned before… Humanity needs to level up and adopt a new “data literate” collective consciousness, but this won’t be easy and it sure as hell won’t happen overnight.
The important thing to remember when faced with such a massive challenge is to start with baby steps and only look at what’s immediately in front of you. Some examples of adopting a more “data literate” mindset would be…
- Put your skeptical glasses on → The next time someone is ranting to you about a recent news headline or sensational idea, ask yourself… Where are these numbers coming from? Are they the whole picture or just a small subset of it?
- Learn the lingo → Next time you come across a word from the worlds of statistics or machine learning takes 2 minutes (no more, no less) to see if you’re able to intuitively understand it. This is a tiny habit, but something that can build over time… The more words you begin to intuitively understand, the more your perspective on the world begins to change (e.g. your “blinders” are lifted).
- Here are a few words you could start with Standard deviation, mean, median, linear regression, and population.
- Ignore the math → Yes… Ignore it! I know it might seem counter-intuitive for me to say this, but in all honesty, the majority of us will not be creating algorithms from scratch or doing academic research. The more important piece is for the rest of humanity to understand the high-level concepts and know-how to look at an opportunity or problem in a “data literate” way. So if you run into a long list of formulas, I want you to IGNORE them and keep moving.
Thank you for joining me on this journey… It’s been an eye-opening one, not just for me, but for all of us.
Diving Deeper
While digging into this paper I’ve come across some reliable resources summarizing and diving deeper into certain areas, check them out below.
- Original Paper – here
- Climate AI Website (created by the team that wrote the paper) – here
- Future Life Institute Podcast Interview – Part 1 here & part 2 here
- Blog and Video Summary – here and here
- Two workshops reviewing the research – here and here