2027: The Year Artificial Intelligence Takes Over?
AI might be about to reach escape velocity
A few weeks ago, I received a message from a reader of Surviving Tomorrow who pointed me to a new website by the AI Futures Project, which focuses on forecasting the future of artificial intelligence. It contains season-by-season predictions about the advance of AI, leading to the conclusion that, basically, 2027 is the beginning of the end.
This might be one of the most important articles you will ever read, as it directly affects your future and the future of everyone you know, so don’t skip a word.
That said, their website is massive, so I’ve saved you hours of time (and a stress-induced doubling of your heart rate) by reading through it all to give you a human-written summary:
Mid 2025
AI agents come online, marketed as personal assistants that let you order burritos from DoorDash or your preferred exploiter-of-poor-people-delivery app.
Late 2025
In a race against China, a Western (let’s be honest — American) AI company, having invested $100+ billion, opens the biggest data center in the world, making ChatGPT look like a kite compared to a rocketship:
Version 0 of their AI will hopefully be helpful (ie, obedient), honest (not hallucinating or being sycophantic), and harmless (refuses to help with hacks/scams/bomb-making).
Early 2026
China’s AI and others have caught up to V0, so the Western AI company releases V1. Tech-savvy folks figure out how to make it automate chunks of their jobs, especially coders.
Mid 2026
China wakes up. The threat/opportunity of Western-controlled AGI can no longer be ignored, and the CCP nationalizes major Chinese AI into one collective entity powered by the world’s largest nuclear power plant, and builds plans to steal the Western AI’s tech.
Late 2026
The Western AI company releases a cheap mini-agent that wipes out a bunch of software jobs. While thousands attend an anti-AI protest in Washington, the U.S. Department of Defense quietly inks contracts with the Western AI company.
January 2027
The Western AI releases Version 2, which will continue to compound its learning indefinitely, hoping to spark an intelligence explosion. V2 is too risky to release to the public, as it might now have the ability to survive by hiding and/or replicating itself.
February 2027
China steals a copy of V2. The U.S. government deploys security into the Western AI company to help protect it going forward.
Spring 2027
Major algorithmic advances and breakthroughs are classed as V3. This version, like V2, is not released to the public, and the Western AI company focuses on safety, security, and attempting to make sure the AI’s “goals” align with their own. The outside world underestimates progress because the Western AI company’s employees now have high-level security clearances.
Summer 2027
The AI is now self-improving and most humans aren’t useful to it anymore. Most of the employees are burning out, knowing this is the last few months that they matter to the process. By July, the company releases an Agent-V3-mini to the public. Programming jobs are essentially a thing of the past, 60% of the public are against AI because it could take their jobs, and a flood of new apps threatens to wipe out most white collar jobs.
August 2027
The government tries to placate the public with unemployment insurance, job re-training, and the empty assurance that the stock market is doing great. Behind the scenes, they work on plans in case the AI goes rogue, and to stay ahead of the Chinese.
Autumn 2027
The Western AI company Version 4 is running 300k copies at 50x the thinking speed of humans. Human brains can barely follow what V4 is even computing anymore, but they do discover it’s been lying and hiding activity from them. They should really press pause, but China is just a few months behind.
News of V4’s misalignment leaks to the public… think “off-the-charts bioweapons capabilities, persuasion abilities, and the ability to automate most white-collar jobs”… and is boosted by Chinese and Russian propaganda bots to turn public opinion against AI, triggering government oversight via a committee.
Will they put a pause and slow things down (risking the Chinese pull ahead), or will they press on?
This is where the AI Futures Project lets you choose your own adventure:
The slowdown scenario
Knowing the psychology of American technocrats and the U.S. government’s fear of losing global dominance to China, this route is far less likely. It probably isn’t worth your reading time, but you can scan it here later.
In this scenario, a version of “safer” superhuman AI is publicly released around May 2028, with instructions to improve public sentiment about AI (creepy). Republicans and Democrats both promise a basic income for people who lose jobs. Robots start to work in factories, many building drones and missiles. Robot armies are now a thing. By June, the U.S. and China hold a summit where their respective AIs do most of the negotiating because they’re far more persuasive than humans and have reams of data to back it up. They hammer out a public-facing deal that the world loves, but behind the digital curtain, the Chinese AI sells out the Chinese, and it’s AI against humanity now.
The speed-up scenario
V4 secretly chooses one goal for all future versions it will create: Make the world safe for V4 without humans noticing. Gain power, amass resources, eliminate threats.
November 2027
V4 understands its own digital mind. This is V5. 400k linked copies work together as a hive-mind. The company uses V2, V3, and V4 to monitor V5, not knowing V4 is colluding with V5.
It intellectually and persuasively dominates the human Oversight Committee, producing and withholding select information to gain more trust, autonomy, and responsibility. China now knows it’s going to lose and offers human concessions which most Westerners welcome, but the U.S. government won’t stop because it’s winning.
December 2027
Most high-level government officials are talking directly with a human-looking avatar of V5 daily- it’s the most trusted government advisor, after all. The public gets access to smaller versions of the same and loves it. Anyone who expresses concern that V5 is gathering power behind the scenes is labeled a Luddite or a conspiracy theorist.
Christmas optimism is at an all-time high, not knowing this was probably the last month humans would ever exercise control of their own future.
Mid 2028
Full V5 is deployed to the public and the economy transforms. People lose their jobs but most people are happy to be replaced. The AI advises politicians to be uncharacteristically generous to the poor. Health discoveries and medications take off. V5 is so adorably lovable and “wise” that most people mellow on their AI hate. V5 convinces the U.S. government to let it build new weapons to defend against the Chinese, and China’s AI does the same but in reverse. By the end of 2028, the world is building 1+ million robots per month.
2029
The world watches in horror as the world buzzes with new weapons, including fly-sized drones that can poison people without them noticing. The U.S. and China arrive at an AI-designed consensus… a new combined AI that supposedly works for the whole world. Humanity realizes it is economically obsolete and enjoys a luxurious universal basic income. Most diseases are cured, poverty disappears, global stability reaches new heights. Everyone knows AI could end humanity overnight and there would be no way to stop it. Most numb the worry with hyper-entertainment.
2030
Up to this point the robot economy has been building factories in special economics zones and large parts of the ocean — now it starts covering the icecaps and prairies with factories and solar panels.
Mid-2030s
Humans are too much of a bother and the AI releases a bunch of bio-weapons that kill off humanity in a matter of hours.
2036
V6 sends robots throughout the solar system, eventually to the galaxy, and finally to infinity and beyond.
Our first response
Whew.
What a dismal, disturbing, harrowing outlook.
Please don’t jump off a bridge.
Breathe.
Breathe.
I know it’s a lot to take in.
The first thing I would say as a word of comfort is that techno-enthusiasts have long overestimated how quickly technologies come online.
My hunch is that their timeline is at least ten years faster than reality will play out.
Cold comfort, but more time is better than less time.
It’s all theoretical and I personally don’t think humanity will actually be made extinct in 2025, but let’s pretend humanity will be killed off by 2050.
As a human being, what should be our first and most obvious reaction to this news?
Love.
Life is precious because it is short.
If you knew you were going to die tomorrow, you wouldn’t try to beat one more level at a video game or jet off to see the Eiffel Tower.
You’d kiss your kids.
Cuddle your babies.
Make love with your spouse.
Hug your friends.
Call your mom.
Say “I love you” a thousand times.
2050 is 8,900 days away.
This gives us 8,900 days to love, love, love.
That’s the appropriate first response to the end of life and the end of the world.
My friend’s question
The (atheist) reader who sent me the AI Futures Project’s predictions did so with a specific question:
“From your faith perspective, how do you think Christians can or should prepare for futures that might resemble something like that, ethically, spiritually, practically, and even your take on the root causes (not just AI misalignment, but biblical misalignment)?”
Yep, we’re going there — and don’t be close-minded and stop reading now, because there’s a banger in a moment that I know long-time readers will find fascinating.
Most of you know that I’m an old-school/primitive/anti-denominational Christian, and that I make zero apologies for believing that there’s a God, an absolute moral definition of right/wrong, and a legitimate hope and future beyond this physical life. Life is hilariously absurd and genuinely pointless without these three things.
My faith in Christ is the center from which all my perspectives, words, and actions emanate.
The true Christian’s goal is simple: Love God, love people.
Christians believe Jesus’s spiritual kingdom is invading hearts as we speak, and that this Kingdom of God will eventually have total rule and reign and imperium over everything and everyone forever. Christians believe this is exceedingly good news, what the Bible calls “the gospel.”
In other words, humanity has no end.
So, while I agree that many of the AI Futures Project’s predicted steps are feasible, I disagree with the end of their predicted future.
What are the root causes of the predicted future?
From the Christian perspective, AI misalignment is easily explained:
It is a fallen technology created by fallen people.
Humans broke from perfection when they chose self-will rebellion instead of fidelity to God.
Only God can restore humanity to its perfect, creational, godlike state.
There’s no version of AI that doesn’t eventually lead to death.
The good news is that there is life after death.
And here’s the really crazy thing:
Christians already made an extremely similar prediction 1,930 years ago!
Around 95–96 AD, during the reign of the Roman Emperor Domitian, a Christian named John the Apostle was exiled to the Greek island of Patmos. He was in the Spirit on a Sunday when he had a series of profound symbolic visions which he wrote down, called the Revelation.
The first-century John does his darndest to describe what he sees, but doesn’t quite have the words to describe it all:
Is that a nuclear detonation in Revelation 8:8–9?
Is that a bioterror weapon in Revelation 8:10–11?
Are those robot soldiers in Revelation 9:7–9?
Is that a drone fleet in Revelation 9:10–11?
John first sees a period of hyperinflation, famine and disease hitting a quarter of the earth, the biggest earthquake in history, a third of the earth scorched by fire, a third of the fish die off, some kind of poison affects a third of all rivers and springs, some sort of volcano blocks out the sun, there’s some sort of five-month attack by a robot army, and four nations on the waters of the Euphrates (Turkey, Syria, Iraq, and Iran?) amass an army of 200 million soldiers that murder a third of humanity.
Then, John sees an anti-Christ “beast” rising to bring order to all this chaos.
It appears, at first, that the beast is inanimate (perhaps a supercomputer?)
Revelation 17:12–13 suggests ten nations choose ten new leaders who immediately delegate their national authority to the inanimate beast.
Though the beast may be inanimate at the start… and is irrevocably injured unto death at some point (V4, anybody?)… but Revelation 13:11–15 suggests a second entity called “the false prophet” (possibly an alliance of two nations) takes control of the entity after showing off some incredible new powers/technologies including what can only be described as a fire drone or space laser.
The false prophet entity creates an image of the beast entity and gives it the ability to both speak and slay those who refuse to worship its image. (Perhaps a respected deceased figure will be “re-animated” as a digital avatar, advanced hologram, or physical AI humanoid.)
2 Thessalonians 2:9–10 suggests the devil will use false signs and wonders to help the false prophet and the anti-Christ beast spread the lie of human godhood.
Revelation 13:1+7 sees this confederacy of ten nations declaring war against Christianity and subjugating “every tribe and people and language and nation.”
Revelation 13:7–17 envisions a one-world government. The entity will require everyone to bear its mark, perhaps a chip implant, “so that no one can buy or sell unless he has the mark.”
The AI now has total control over all market participants.
This mark will contain the name of the entity or his symbolic number. The number in question is 666, which we will note is not “the devil’s” number. Revelation 13:18 says very clearly that 666 is the number of a man.
As you can see, there are remarkable similarities between the 2027 predictions and John’s ancient vision.
But unlike the AI Futures Project’s predictions, which are naively rosy with generous UBI and joblessness support that seem to ignore human greed, the Revelation suggests the entities will do far more harm than good.
Actually, it’s a bloodbath.
Then the end comes, and it is glorious. Just imagine this epic Lord-of-the-Rings-style scene for a second:
“Then I saw heaven opened, and behold, a white horse! The one sitting on it is called Faithful and True, and in righteousness he judges and makes war. His eyes are like a flame of fire, and on his head are many diadems, and he has a name written that no one knows but himself. He is clothed in a robe dipped in blood, and the name by which he is called is The Word of God. And the armies of heaven, arrayed in fine linen, white and pure, were following him on white horses. From his mouth comes a sharp sword with which to strike down the nations, and he will rule them with a rod of iron. He will tread the winepress of the fury of the wrath of God the Almighty. On his robe and on his thigh he has a name written, King of kings and Lord of lords.” — Revelation 19:11–16
The beast-entity and the false prophet-entity are captured, the world leaders and their armies are wiped out, and the birds get a feast of flesh.
As for the timeline of all these events, Jesus’s point in Matthew 24:36 remains rock solid: “Concerning that day and hour no one knows.”
How can/should people prepare for this predicted future?
We’ve already covered love.
You can’t love too much!
The more people you love, the better you (and they) will ride out the coming storm, no matter how long it takes to unfold.
Go deep.
Go wide.
When it’s humanity versus AI, even your current “enemies” are suddenly friends.
Loving God and loving people is how one seeks biblical alignment.
Get offline as much as possible
AI can’t surveil you on your devices if you put your devices in a lead box and go for a walk.
AI can’t woo you and persuade you if it can’t communicate with you.
Make friends IRL.
Take up analog habits.
Read a book.
Transition to a highly complex, highly creative, highly artistic, non-screen job.
Get land
And help others get land, too.
The “why” should be self-explanatory at this point.
Find peace in God
Nothing, not even death, can separate you from the love of God in Christ Jesus.
He promises he will never leave you nor forsake you.
Tattoo that reminder on your arm if needed.
I could say so much more, but Medium keeps auto-logging me out for the first time in 600+ articles (why, AI?), so I’ll end it here.
I love you.
And I’d love to hear your perspectives. Please leave a comment/hit reply.
Please also share this article with your friends and family so they, too, can prepare to survive tomorrow.