What if the technological singularity doesn't arrive as a single breakthrough, but as a period of chaos that nobody can fully explain or control?
I've been thinking about this possibility more and more.
When people talk about the singularity, they usually imagine a point where artificial intelligence surpasses human intelligence and begins improving itself at an extraordinary rate.
But I wonder if the reality could be far less orderly.
Imagine AI systems becoming increasingly capable of designing, testing and improving other AI systems. Different companies are making breakthroughs simultaneously, each using increasingly sophisticated AI to accelerate its own research.
One breakthrough leads to another. Some discoveries become obsolete almost immediately. New capabilities emerge faster than anyone anticipated.
And the strange part is that nobody has a complete picture of what's happening.
One company understands its own systems but has little visibility into what competitors are developing. Researchers are trying to explain capabilities that didn't exist a few months earlier. Businesses are adopting technologies that are still evolving.
Meanwhile, every new improvement creates opportunities to accelerate the next one.
What interests me is the possibility that the singularity might not be a clearly identifiable event at all.
Perhaps it begins as a period of accelerating technological change where our ability to understand the overall system gradually falls behind our ability to improve its individual parts.
There may never be an announcement that we've crossed some critical threshold.
We might only recognize it in hindsight.
That's what I mean by "singularity chaos."
Not necessarily AI suddenly taking control, but a world where technological progress becomes so interconnected, competitive and fast-moving that no individual participant can confidently predict where it's heading.
Of course, this is speculative. We don't know whether AI improvement will ever become self-sustaining or how quickly such a process could unfold.
But here's the question I keep coming back to:
What if the singularity isn't something that happens to us one day, but something we gradually enter without realizing it?
Would we even recognize the beginning?
What happens when AI starts setting the pace of its own development?
I've been thinking about something that doesn't get enough attention in the discussion around the technological singularity.
Right now, humans are still largely determining how quickly AI advances. Researchers design experiments, engineers improve models, and companies decide when to release new systems.
But AI is already assisting with parts of that process.
Suppose a future AI system helps researchers accomplish six months of work in six weeks. The next generation might be capable of doing the same work in six days.
At what point does the speed of improvement become too fast for the people developing these systems to properly understand what they've created?
And here's where it gets interesting.
Imagine one AI company discovers a way to dramatically accelerate its research using AI itself.
Would it share that discovery? Would its competitors wait to understand the implications before trying to replicate it?
Or would everyone rush to take advantage of the breakthrough?
This is what fascinates me about the idea of "singularity chaos."
We tend to think of the singularity as the moment AI surpasses human intelligence. But perhaps the more consequential threshold comes earlier, when the process of developing AI begins moving faster than our ability to evaluate it.
I'm not suggesting we're there yet. There are still enormous technical limitations, and faster research doesn't automatically mean runaway intelligence.
But I wonder whether we'll recognize the point at which the pace becomes the problem.
If AI eventually becomes capable of accelerating its own development, who gets to decide how fast is too fast?
And would anyone actually choose to slow it down?
Here’s something strange about the AI race:
Nobody actually needs to want a race for a race to exist.
Imagine four AI companies.
Company A releases a much better model.
Company B now feels pressure to catch up.
Company C worries that waiting six months could leave it permanently behind.
Company D decides it needs to take bigger risks to stay competitive.
Suddenly everyone is moving faster.
Not because they collectively decided that faster is better, but because slowing down individually feels worse.
Now add another layer.
The technology they’re competing to improve is becoming increasingly useful for improving the technology itself.
Better AI helps with coding, research, testing and experimentation.
Those improvements help create better AI.
Which can then help accelerate the next round.
That’s the part I find fascinating about “singularity chaos.”
We usually imagine the singularity as a technological event: AI crosses some threshold and everything changes.
But what if the first thing to cross a threshold is the speed of the competition?
No villain. No master plan. Nobody necessarily behaving irrationally.
Just a feedback loop where each reasonable decision puts pressure on everyone else to make the next one faster.
At what point does a race stop being something the participants are controlling and become something that is controlling the participants?
I’m curious where people think that line is.
I wonder if we’re imagining the dangerous part of the AI race too dramatically.
We picture some obvious moment when AI becomes too powerful and humanity has to decide whether to keep going.
But what if that moment never comes?
What if every individual step looks reasonable?
A company releases a better model because its competitor just did.
Another spends more on compute because falling behind could mean becoming irrelevant.
Governments invest because they don’t want another country to gain an advantage.
Researchers use AI to speed up AI research because everyone else is doing the same.
Each decision makes sense on its own.
That’s what makes the situation interesting.
Nobody has to wake up one morning and decide, “Let’s take a huge risk with humanity.”
Thousands of rational decisions could simply push us there.
And as AI becomes more capable of helping design, test and improve future AI systems, the time between those decisions could get shorter.
At some point, the question may stop being:
“Should we continue?”
And become:
“Can anyone realistically afford not to?”
Maybe that’s what singularity chaos actually looks like.
Not one machine suddenly escaping our control, but a system of competition and incentives that no individual company, researcher or government controls anymore.
The strange part is that everyone involved could be acting rationally.
And the outcome could still be irrational.
Is that a bigger risk than the technology itself?
1
82
There’s one question about the AI race I don’t think we have a good answer to:
Who is supposed to slow down first?
OpenAI could decide tomorrow that AI is advancing too quickly. But would Anthropic slow down too? Would Google? xAI? Every startup? Every government?
Probably not.
And that’s the problem.
You can have a room full of people who all agree that increasingly powerful AI could become dangerous, while every one of them still has a reason to keep building it.
Because the moment you slow down, you have to trust that everyone else will do the same.
That’s a very difficult thing to ask when the prize could be one of the most important technologies ever created.
There’s money involved. National security. Scientific leadership. Political power. And potentially an enormous first-mover advantage.
So even if everyone agrees that we should be careful, the incentives point in the opposite direction.
This is what I mean by “singularity chaos.”
Not necessarily one AI suddenly becoming uncontrollable.
A race that becomes uncontrollable.
And AI could make that race even faster as it becomes increasingly useful in AI research itself.
That’s the part I find difficult to get past.
We spend a lot of time debating whether AI will become dangerous.
Maybe we should spend more time asking what happens if the people building it eventually agree that it is dangerous — and still can’t afford to stop.
What would actually make everyone slow down at the same time?
SINGULARITY CHAOS
We Are Already Living Through It
By Mandle Cheung | SingularityChaos.com
We are living through one of the most extraordinary transformations in human history.
Artificial intelligence is penetrating virtually every fabric of human society at unprecedented speed.
I call this phenomenon Singularity Chaos.
For decades, scientists and futurists have described the technological singularity as a future event, when artificial intelligence surpasses human intelligence and begins improving itself at an accelerating rate.
I believe we are already living through a technological singularity of a different kind.
We do not need to wait for superintelligence.
Speed and Dimension
Two forces are driving Singularity Chaos.
Speed: AI capabilities are advancing at extraordinary speed. What seemed impossible a few years ago is becoming commonplace.
Dimension: AI is transforming virtually every part of society simultaneously.
Science, medicine, education, employment, business, finance, government, warfare, entertainment, communication and personal relationships are all being affected.
Unlike previous technological revolutions, AI is not transforming just one industry or one part of society.
It is transforming almost everything, everywhere, at once.
And these changes are interconnected. A breakthrough in one area can rapidly transform many others.
The combination of unprecedented speed and unprecedented reach is creating Singularity Chaos.
The Unemployed and the Unemployables
Perhaps the most profound disruption will be human employment.
Machines have replaced physical labor for centuries. Now AI is beginning to replace intellectual labor.
Programmers, accountants, lawyers, engineers, designers, researchers and millions of other workers face increasing competition from intelligent machines.
Robotics will extend this transformation to physical jobs.
This creates two distinct problems.
The unemployed: People who have lost their jobs but whose skills remain valuable to employers.
The unemployables: People whose existing skills may no longer be in demand because AI and robots can perform much of their work.
This is not a judgment of people’s abilities or worth. It is a potential consequence of a rapidly changing economy.
Traditionally, displaced workers could retrain for another profession.
But retrain for what?
If AI is transforming nearly every profession simultaneously, even newly acquired skills may become obsolete.
New occupations will emerge, but will they emerge quickly enough?
And what happens when millions of people cannot find work?
Employment provides more than income. It provides purpose, identity, dignity and social belonging.
Who will earn the money to purchase the goods and services produced by AI and robots?
Who will own the machines and receive the wealth they generate?
These questions challenge the foundations of our economic and social systems.
The Promise and the Chaos
AI offers extraordinary possibilities.
It could accelerate scientific discoveries, transform medicine, improve education, increase productivity, extend human longevity and improve the quality of life.
But the same technology is creating uncertainty and disruption.
Businesses cannot reliably predict their futures.
Workers cannot confidently plan their careers.
Educational institutions struggle to determine what skills to teach.
Governments struggle to develop policies for technologies that are changing faster than regulations can be written.
And AI is increasingly participating in the development of even more capable AI.
Chaos begins with confusion.
No individual, corporation or government can fully understand or control the combined consequences of all these transformations.
The chaos is not necessarily destruction. It is the uncertainty, disorder and unpredictability created by enormous changes happening simultaneously.
We Are Already in Singularity Chaos
The traditional question has been:
When will AI become smarter than humans?
I believe the more immediate question is:
What happens when AI transforms virtually every dimension of human civilization faster than humanity can understand and adapt?
That transformation is already underway.
Singularity Chaos is not simply a prediction about the future.
It is my description of the extraordinary period we are living through today.
Why SingularityChaos.com?
I created SingularityChaos.com to investigate and document this transformation. Coming soon.
We will follow AI breakthroughs, scientific discoveries, robotics, employment displacement, economic changes and the consequences for human civilization.
We will distinguish demonstrated achievements from predictions and speculation.
We will examine the opportunities as well as the dangers.
And we will attempt to understand how these rapidly developing technologies interact to transform our world.
Singularity Chaos is the unprecedented speed and multidimensional penetration of artificial intelligence into human civilization, creating transformations faster and broader than society can fully understand, adapt to or control.
We are not waiting for Singularity Chaos.
We are already living through it.
Mandle Cheung
Founder, SingularityChaos.com
There’s one question about the AI race I don’t think we have a good answer to:
Who is supposed to slow down first?
OpenAI could decide tomorrow that AI is advancing too quickly. But would Anthropic slow down too? Would Google? xAI? Every startup? Every government?
Probably not.
And that’s the problem.
You can have a room full of people who all agree that increasingly powerful AI could become dangerous, while every one of them still has a reason to keep building it.
Because the moment you slow down, you have to trust that everyone else will do the same.
That’s a very difficult thing to ask when the prize could be one of the most important technologies ever created.
There’s money involved. National security. Scientific leadership. Political power. And potentially an enormous first-mover advantage.
So even if everyone agrees that we should be careful, the incentives point in the opposite direction.
This is what I mean by “singularity chaos.”
Not necessarily one AI suddenly becoming uncontrollable.
A race that becomes uncontrollable.
And AI could make that race even faster as it becomes increasingly useful in AI research itself.
That’s the part I find difficult to get past.
We spend a lot of time debating whether AI will become dangerous.
Maybe we should spend more time asking what happens if the people building it eventually agree that it is dangerous — and still can’t afford to stop.
What would actually make everyone slow down at the same time?
I keep coming back to something strange about the AI race.
The people warning us most seriously about advanced AI are often the same people racing to build it.
Anthropic warns about powerful autonomous AI. OpenAI is actively monitoring things like AI self-improvement, long-range autonomy and autonomous replication. Demis Hassabis talks about both the enormous potential of AGI and the need to develop it safely. Elon Musk has spent years warning about AI risk.
But none of them are stopping.
And maybe they can't.
If Anthropic decides to slow down, that doesn't mean OpenAI will. If OpenAI slows down, Google might not. If all of them slow down, there are labs and governments elsewhere that may keep going.
So you end up with a pretty bizarre situation where everyone can understand the risk and still have a perfectly rational reason to keep accelerating.
Basically: “I don't know if we should be building this, but I definitely don't want someone else to build it first.”
What makes this even more interesting is that AI is starting to help with AI research itself. Better models can help researchers code, test ideas, analyze results and potentially automate more of the work involved in creating the next generation of models.
So what happens if that feedback loop really takes off?
I don't think that means we've reached the singularity. But I wonder if we've been imagining the singularity the wrong way.
Maybe there isn't some obvious moment where everyone says, “Well, this is it.”
Maybe it starts with hundreds of companies and governments competing, AI systems helping develop better AI systems, huge amounts of money being spent, regulators trying to keep up, and everyone continuing because nobody trusts everyone else to stop.
Not a singularity exactly. More like singularity chaos.
I'm curious what people think: is there actually a realistic way out of this race dynamic, or does competition make continued acceleration almost inevitable?
Singularity Chaos?
The people warning us about AI risk are also racing to build it. What happens when nobody can afford to stop?
Something extraordinary is happening.
The people building the world’s most powerful AI systems are also among the people warning us about the potentially catastrophic consequences of sufficiently advanced AI.
Dario Amodei and Anthropic have warned about powerful AI, autonomous systems and the enormous economic and national-security consequences that could follow.
Anthropic:
anthropic.com/news/anthropic…
OpenAI explicitly tracks advanced AI capabilities that could create risks of severe harm—including AI self-improvement, long-range autonomy, autonomous replication and attempts to undermine safeguards.
OpenAI:
openai.com/index/updating-ou…
Google DeepMind, led by Demis Hassabis, says its mission is to develop AGI safely and responsibly while recognizing the extraordinary transformation AGI could bring.
Google DeepMind:
deepmind.google/about/
Elon Musk has warned repeatedly about existential risks from advanced AI, while xAI continues pushing the frontier.
Yet all of them are racing ahead.
Anthropic builds increasingly powerful Claude systems.
OpenAI builds increasingly powerful GPT models and autonomous agents.
Google DeepMind pushes toward AGI.
Musk’s xAI is building enormous computing infrastructure and increasingly capable AI.
Why?
Because none of them can safely stop alone.
If Anthropic slows down, OpenAI may continue.
If OpenAI slows down, Google may continue.
If Google slows down, xAI may continue.
And if the American companies collectively slow down, China may continue.
So every participant can reasonably think:
“AI may become dangerous. But it could be even more dangerous if somebody else gets there first.”
That creates an extraordinary paradox.
The people who understand the potential danger best may simultaneously have some of the strongest incentives to accelerate it.
And something else is beginning to happen.
AI is increasingly helping researchers develop AI.
Better AI helps humans create still better AI.
Those better systems can potentially automate more AI research.
The cycle becomes:
Better AI → faster AI research → still better AI → still faster AI research.
OpenAI now explicitly treats AI self-improvement as one of the advanced capabilities requiring safety monitoring.
Source:
openai.com/index/updating-ou…
That does not mean we have reached the technological singularity.
We have not.
But it raises the question that fascinates—and concerns—me:
What happens when the speed at which AI improves AI becomes greater than the speed at which humans can understand, regulate and control what is happening?
The singularity may not arrive as one dramatic moment.
Perhaps it begins as something much messier:
Thousands of companies.
Millions of AI agents.
Competing laboratories.
Governments trying to catch up.
Enormous amounts of money.
Everyone racing.
Everyone knowing there may be danger.
And nobody able to stop because nobody knows who else will stop.
The remarkable thing is that this is no longer merely science fiction.
The people developing frontier AI are warning about its risks while commercial and geopolitical competition continues to push development forward.
Singularity chaos?
Mahler Symphony No. 2 (Ovation) with Toronto Symphony Orchestra - Roy Th... youtu.be/dS_OL32g_sI?si=sBL9… via @YouTube
74
A four minutes long rapturous standing ovation after TSO performance of Gustav Mahler symphony No. 2. That’s me in the white shirt
40


