Quarterly Update Q1 2023
News sometimes influences the topics we cover in Uncharted Territories. I gather it all and share with you once every quarter.
Here are this quarter’s updates:
Uncharted Territories Is Growing
When Will the Singularity Arrive?
Is Remote Work Here to Stay?
German Energy Independence
German Industry Is Saved
Should We All Learn English?
The 36 Questions that Lead to Love Are Different
[Premium members only below this]
Egyptian Density Visualizations
Ethiopia’s Civil War Has Ended
Punt Was in Eritrea
Money Also Prevents Us from Building Up
Global Taxation
How the Doggerland Disappeared
We’re Like Neurons
Why Is Fertility Down?
More Bayesian Precision
Uncharted Territories Is Growing
We just passed 50k readers!
The reason has been the GeoHistory Twitter threads, which together have gotten more views than The Hammer and the Dance. Here are the main ones:
When Will the Singularity Arrive?
Experts believe the world as we know it will end in 7-18 years.
The singularity is the moment when AI becomes so good that it reaches a God level of intelligence. After that, what will happen? All bets are off, because by definition we can’t know what a God-like intelligence can think. So when will that moment arrive? Let’s see what experts think.
For this type of question, the best source is prediction markets. And Metaculus, one of the best ones, thinks it’s coming pretty soon.
We can divide the path to superintelligence into three steps:
Weak general AI: An AI that can use general training to pass tests better than most humans, without specific training on that test. For example, have a better score than 75% of students on an SAT test with less than 10 examples of previous SAT tests.
Artificial General Intelligence (AGI): An AI that can do most tasks better than most people.
Superintelligence: An AI whose intelligence is so over that of humans that we can’t comprehend it.
When do experts think each one will arrive?
People used to think that weak AGI would arrive in the mid-to-late 2040s. But this year, that has shrunk to… 2027! And then they think that the time between weak AGI and superintelligence will be 40 months, or just over 3 years.
Meanwhile, when you ask directly the date of full AGI, they currently believe it will be around 2041.
In other words, the market believes that in 7-18 years, we will have AIs that can do most tasks that humans can do, at a fraction of the cost. And that they will also become much more intelligent than us. This will change absolutely every aspect of life.
The world as we know might be about to end in 7-18 years.
Remote Work
I’m a big believer that remote work is here to stay, so I’m looking at indicators to gauge this. Here’s an interesting one:
According to this, about a third of work in the US that used to be in the office is remote and might stay that way.
There’s still a big gap between supply and demand, where bosses want people to be in the office, but employees don’t want to: Only 15% of job postings offered remote work, but they receive 50% of applications!
I don’t know what you think, but for me, with unemployment at 3.4% in the US, a cataclysm would be needed for employers to gain the upper hand. If the vast majority of talent wants to work remotely, who do you think will get the best people, the companies that accommodate them, or the rest? Who, then, will become more productive? Who will win the market? I’m bullish on remote work.
German Energy Independence
I’ve been arguing hard for Germany to keep its nuclear reactors open and move away from Russian gas. They did keep some nuclear reactors open—although it’s not clear for how long.
What they did achieve, however, was the primary goal: Being free of Russian gas!
The warmest European winter ever might have helped. But this also comes with huge efforts: From replacing gas with alternatives, increasing imports from the Netherlands and Norway (thanks!), and sourcing it from other countries.
But that’s not all. They’re free of all energy from Russia!
Coal, gas, and oil. All of it.
Good job, Germany!
Unfortunately, the downside is that they’ve been reviving their coal industry to do it—which as we know is highly polluting. For example, this village of Lützerath will be demolished to expand a coal mine.
German Industry Is Saved
One of the fears I had in How to Fight Putin, Climate Change, and Your Energy Bills was that an increase in electricity bills in countries exposed to Russia could condemn energy-intensive industries to death, which would be especially sad in Germany since they had a nuclear alternative.
I proposed a few ways that industry could replace gas. In fact, between the mild winter and their initiatives, German industries did reduce their gas consumption. 75% of German gas-consuming companies said they reduced their gas consumption without affecting production, and 40% said they could go further. Another win for Germany!
One of the measures I proposed was heat pumps. And yes, the heat pump market exploded in Germany last year!
If you’re looking for a way to heat your home cheaply, and you’re in a country where the cost of electricity compared to gas is right, heat pumps are a fantastic option for you, even for your home.
Should We All Learn English?
In Should Everybody Learn English? and The Ineluctable Progress of English, I explained why I think it’s a matter of time before we all speak English: Its pace towards becoming the lingua franca1 of the world is unstoppable because of its network effects.
Now, with remote work, the network effects are even stronger. A person who speaks English can work for an English-speaking company and earn 10x more.
If that’s not a massive incentive for anybody to learn English fast, I don’t know what is.
Conversely, I said only one thing could stop this trend: That AI gets good enough for universal, seamless, instantaneous, perfect translation: If a Chinese, a Brazilian, and an American are speaking their own languages but they all hear the others in their mother tongue, learning English might be too costly.
Some new tools suggest this might be likely.
For example, EzDubs has a tool for automatic voice translation of any video. I tried it with a video that an Australian TV channel made of my India thread.
Here’s the Spanish translation. It’s acceptable, and this is just the very beginning.
And of course, there’s Google Translate, which is still surprisingly good—despite Google taking too long to ship now.
Meanwhile, OpenAI has open-sourced Whisper, one of the best transcription and translation models available today. It feels like it’s waiting for someone to make a good product out of it. One of my goals is to play with this type of tool in the future to see how good it is and if I can use it for Uncharted Territories: Many of you have asked for audio versions of my articles, and I think translating them into other languages would be valuable too.
So which one will win? Will remote work make people learn English faster than Generative AI makes it useless? I have no idea. What do you think?
The 36 Questions that Lead to Love Are Different
I like hacking mental processes, so I’ve referenced the concept of the 36 questions that lead to love a few times. It’s a quick, 45 minute session where two people can quickly create intimacy by asking 36 questions that become more and more intimate over time.
Yes, these questions do increase intimacy fast. But there was a secret hidden in the details that Ivan Vendrov discovered:
It turns out that the original research had people cover 40 questions for 1.5h, not 36 questions for 45m. And some of the differences were crucial. Here’s the 5th question:
If you were going to have a personal relationship with your partner, please share what would be important for him or her to know.
Pretty intense.
Then close to the end, four questions escalated things to steam:
The good thing about this is it gives you a good insight into how to create strong emotions quickly: Ask questions that let people be vulnerable, that let them play about you, and then make them imagine the situation you want them to be in.
The bad thing is that it’s less directly applicable to other contexts like business. You’re not going to look a potential business partner in the eyes for three minutes in silence, are you?
What you can, however, is to ask them to imagine scenarios
Imagine a scenario where we’re working together and you’re extremely happy. What happened?
Let’s role play a time when we’re working together and we just heard news that we succeeded tremendously.
Close your eyes and tell me everything you like about us.
If you’re going into serious business relationships like co-founding a company, these might be worthwhile exercises.
Keep reading with a 7-day free trial
Subscribe to Uncharted Territories to keep reading this post and get 7 days of free access to the full post archives.