Demis Hassabis's Stunning Prediction: Less Than 2000 Days Left for the Old World

During a conversation on the Stanford campus, Demis Hassabis made three significant statements.

The first one silenced the entire audience for two seconds.

The second one made everyone start counting the days.

The third statement is actually highly relevant to every entrepreneur.

This article will break down these three statements thoroughly, then highlight 3 common mistakes that Chinese entrepreneurs are prone to making, and finally provide 4 specific actionable suggestions.

We hope today's content will be inspiring for you.

1. Demis Hassabis's 3 Statements, Each Deserves In-Depth Exploration

The first statement: "Humanity is standing at the foot of the singularity."

Notice his choice of words: "the foot of the mountain" — not "the top of the mountain" or "halfway up the mountain." If he thought the singularity had already arrived, he would say, "We are crossing the ridge." If he thought it was still far away, he would say, "There is a mountain in the distance."

"At the foot of the mountain" means: We have already reached the base of the mountain, haven't started climbing yet, but you can already see the slope getting steeper as you look up.

This is not just a metaphor. Demis Hassabis is someone who treats precision as a belief — he is a chess grandmaster, a PhD in neuroscience, and a Nobel Prize in Chemistry laureate. He chose the phrase "at the foot of the mountain" after careful consideration.

The reason he used the term "singularity" is that he sees AGI (Artificial General Intelligence, hereinafter referred to as AGI) being only 4 to 5 years away from us.

In the field of artificial intelligence, "singularity" specifically refers to a turning point in technological development. After reaching the singularity, the intelligence level of technologies like artificial intelligence will surpass human intelligence, entering a stage of "intelligence explosion."

Subsequently, technological growth will become uncontrollable and irreversible, and human society will face fundamental transformations.

The second statement: "AGI will arrive around 2030, with a margin of error of no more than one or two years."

This statement is extremely information-dense, and can be broken down into three layers:

The first layer: He is talking about AGI, not "stronger AI." AGI means that machines can reach or exceed human performance on the vast majority of cognitive tasks — this is a qualitative leap.

The second layer: He has given a time anchor: 2030. That is roughly 3 to 4 years from now. As a top figure in the global AI industry, this is a timeline he is willing to publicly bet on after reviewing training data, computing power curves, and model architectures.

The third layer: "A margin of error of one year." This is far bolder than saying "around 2030" — he is narrowing the range of uncertainty.

The third statement: "The most important thing is to firmly grasp your own initiative. The script for the future has not been written yet. Do not listen to anyone who says the future is already predetermined."

This statement sounds like motivational fluff, but it is absolutely not.

Think about it: Demis Hassabis is standing "at the foot of the singularity mountain," just told you AGI could arrive in 3-4 years, and he knows better than anyone how powerful AI is. So he suggests that everyone embrace and explore the capabilities of AI tools, but what he emphasizes even more is seizing your own agency.

He placed this statement at the end of the entire conversation. It was not a casual remark — it is likely a conclusion he thought about for a very long time.

How much impact do these two things — "at the foot of the singularity mountain" and "AGI arriving around 2030" — actually have?

Let's first thoroughly discuss these two points, otherwise, the weight of the third statement will not land properly.

2. The Picture Painted by the First Two Statements

1. What does 10x impact multiplied by 10x speed equal?

Demis Hassabis said: "The impact (of AI) is ten times that of the Industrial Revolution, and the speed is also ten times. It will happen within 10 years, not 100 years."

What did the Industrial Revolution change?

Production methods shifted from manual labor to machinery;

Urban forms transformed from villages and towns to metropolises; class structures changed from feudal systems to capitalist systems;

The global power landscape shifted from agricultural empires to industrial powers. It reshaped human society.

An impact 10 times greater than the Industrial Revolution means that earth-shaking changes will take place in human society.

What does a 10x speed mean?

The British Industrial Revolution started in the 1760s and was roughly completed by the mid-19th century, taking nearly 100 years.

Now, 100 years are compressed into 10 years — this speed of change will leave most people breathless.

A British textile mill owner, from first hearing about the steam engine to having their entire production line powered by steam, would take roughly 15-20 years. They had an entire generation's time to observe, trial-and-error, and adjust direction.

Now, you do not have a generation's time — you probably only have a two- to three-year window.

This is the true meaning of "at the foot of the singularity mountain": The magnitude of change brought by AI is on the level of the Industrial Revolution, but the time left for you to react is only one-tenth of what people had during the Industrial Revolution era.

Demis Hassabis dares to make this judgment not out of thin air.

He saw AlphaFold solve the 50-year-old protein folding problem that had puzzled the biology community in just a few months; he saw AI devouring work that used to take human teams several years to complete in materials science, weather forecasting, and chip design, at a speed visible to the naked eye.

With this judgment, what assumptions should you make as an entrepreneur? Obviously, you should prepare in advance.

Only by preparing early can your efficiency be higher than others, and your team can find the integration point between AI and your own business one step ahead of others.

If you prepare under the assumption that "it's still early," one day you will suddenly find that the entire track has been redone, and you do not have an admission ticket.

History is full of such lessons.

In 2007, when Steve Jobs held up the iPhone, Nokia's executives thought touchscreens were just toys. Four and a half years later, Nokia's mobile phone business was sold to Microsoft.

In 2015, many people had heard the term "artificial intelligence," but very few companies actually started using deep learning to restructure their businesses.

By the end of 2022, when ChatGPT was released, those who had prepared for seven years and those who had just heard the words "large language model" were no longer on the same starting line.

Humans have a systematic bias: we always underestimate the speed of technological change. Not because we are stupid, but because past experience tells us that change always happens slowly.

But this time, change will not happen slowly.

2. From Panic to Action

So, if you accept the two premises of "at the foot of the singularity mountain" and "AGI arriving around 2030," you will face two instinctive reactions:

One is panic: "Five years? I won't have time to do anything."

The other is to give up: "Since AI can do everything anyway, there's no point in me doing anything."

Demis Hassabis's third statement is directed at these two reactions. Panic is useless, because panic is not action.

Giving up is even more useless, because AGI is not the end of the world — it is the rewriting of the old world. And Demis Hassabis tells you that the power of this rewriting is still in your hands. The only thing worth betting on is your agency.

3. What Exactly Is Agency?

First, let's clarify what it is not.

It is not "you have to work hard" — that is a meaningless platitude. It is not "execution capability," which means doing known things well. But in the AGI era, most of the time you don't even know what you should be doing.

So what is it? Break it down, and it consists of three things.

First: You must be able to make choices, not just optimize.

Suppose you are the owner of a restaurant.

Over the past decade, what has been your core competitiveness? It might be good dishes, a great location, and fast table turnover. These capabilities all have one thing in common: they are all answering "how to make existing things better" — that is optimization.

AGI is best at exactly this. It can cut your supply chain costs by 15%, maximize your table turnover rate, and dynamically adjust your menu to the optimal state based on seasons and customer flow. All the experience you spent 10 years figuring out, it can calculate in a month.

But there is one thing it cannot figure out: Should your restaurant switch to takeout only? Should you use pre-prepared dishes? Should you merge with another brand and switch tracks?

Why can't it figure that out? Because the answers to these questions are not "better" — they are "different."

Optimization is working within a framework, while choice is changing the framework. AGI can achieve the ultimate within the framework you give it, but the task of "what framework to give" cannot be done by it — only you can do it.

After AI takes optimization to the extreme, what is the only thing left? It is choice. Choosing what to do is far more valuable than choosing how to do it.

Second: You must actually get involved, not just keep watching.

Many times, a task that you have been tangled up about for three months and haven't started will become completely clear to you in just three days once you start doing it.

This is not because you suddenly became smarter — it is because when you just sit and think, all you are running in your mind is simulations.

Simulations cost nothing, and they do not generate real information. Only when you actually do it, actually invest money, hire people, meet clients, and get rejected, will reality give you feedback.

AGI can help you simulate ten thousand market strategies, but it cannot take that blow for you, nor can it bear the heartache after losing money. But it is precisely that heartache that will raise the accuracy of your next judgment by a level.

This is also why "wait and see" is the most dangerous strategy. People who wait think they are observing, but they are actually missing out. Truly valuable information can only be obtained by doing. People who wait two years before taking action do not start from zero — they start from a negative number, because their competitors have already iterated several times.

In the era of information explosion, information obtained through action is the scarcest of all. It is obtained with real money, not by searching on a search engine.

Third: You must know who you are, instead of letting the market decide for you.

An entrepreneur has two directions: AI+healthcare and AI+finance, and the data supports both. He chose healthcare. Why? You could say it was intuition.

But what is intuition?

It is all the experience, values, failures, preferences, and beliefs about "what is worth doing" that he has accumulated over decades. These things mix together to form an indivisible whole. You cannot write it into a prompt and feed it to a large language model, because even he himself cannot clearly explain what the weight of each layer is.

AGI can help you build a better medical AI, and it can also help you build a better financial AI. But it cannot answer the question "why healthcare?" — because the answer is not in its data, but in this person's life experiences.

What you choose defines who you are.

Combining the three layers into one sentence: The stronger AGI is, the more valuable agency becomes.

Because AGI has made "how to do things" almost free — execution, analysis, optimization are everywhere. But the three things "what to do," "whether to do it," and "who you are" will always have a limited supply.

4. The 3 Most Common Mistakes

Three types of mistakes, each more hidden than the last.

The first one: Treating cost-cutting as a strategy.

For example, laying off customer service staff and replacing them with AI. Response times are reduced, and tickets are closed faster. The boss looks at the reports and thinks it's correct. A few months later, the renewal rate starts to drop. Customers can't tell exactly what's wrong, but they just feel "it's not as close as before."

This way of dying is the most unjust: All visible indicators are rising, but the cause of death is hidden in places you cannot see.