AI Godfather Hinton’s Apocalypse Warning! You must lose your job, and the AI trillion foam can “win” by gambling

AI Trillion Gambling Game, Musk is Getting Richer, But Are Humans collectively Unemployed? Hinton’s latest doomsday warning: tech giants cannot make money without layoffs.

This AI trillion dollar gamble can only be profitable through layoffs.

This is the latest doomsday warning issued by AI godfather Hinton.

Recently, Hinton bluntly warned in an interview with Bloomberg that super intelligent AI is coming, which will not only replace humans, cause mass unemployment, but even threaten human survival.

Tech giants are personally creating ‘alien invaders’ but still addicted to business competition.

Now, the ultimate foam of AI is just around the corner.

Expected next year, Microsoft Meta、 Google and Amazon, the four giants, have spent up to $420 billion ($360 billion this year) in the field of AI.

However, OpenAI alone has signed computing power contracts worth over $1.4 trillion. This large order tightly binds Silicon Valley AI giants to buy GPU training models and then make profits.

The biggest winner behind this AI feast is NVIDIA.

After the GTC 2025 conference, its market value skyrocketed to $5 trillion, and Loop Capital’s latest prediction is that it is expected to exceed $8.5 trillion in the future.

The host asked Hinton a heavy and thought-provoking question: Can trillion dollar investments yield returns without destroying employment?

Hinton’s answer was brief and powerful, ‘To make money, one must replace human labor.’.

This sentence is like a heavy hammer, shattering those who are immersed in the AI competition. In just ten minutes of conversation, Hinton spoke many heartfelt words.

Human survival is not important, Silicon Valley giants only want to win

After leaving Google, Hinton did not continue to develop AI, but instead constantly warned people about the risks of AI.

Four months ago, in an interview with over ten million views, he issued a warning to the world, exposing the dangers of AI that no one can handle.

He said that people’s awareness of risks is indeed deeper than before, but just knowing is not enough – we must take action.

Hinton compares AI to a delayed ‘alien invasion’:

If a telescope discovers an alien invasion fleet that arrives on Earth within ten years, we will definitely panic and make every effort to respond.

The current situation is that we are personally “manufacturing” these extraterrestrial (AI) beings, which will be smarter than us.

Humans must seriously consider how to coexist with them? Coexistence “and” control “are precisely the two issues that Hinton repeatedly pondered.

When it comes to AI security, he provided individual comments on industry giants——

Google and Anthropic place much higher emphasis on security issues. Dario Amodei, Demis Hassabis, and Jeff Dean take security seriously, but they are also involved in fierce business competition.

Meta is a typical representative of irresponsibility.

OpenAI was originally established to ‘responsibly develop AI’, but now it is increasingly deviating from this goal, and the best security researchers are leaving one after another.

Hinton criticized this by saying that many companies verbally promise safety but instead invest resources in commercial competition.

They often say, ‘Don’t worry, we have top scientists watching. The real danger is still early, and our experts will discover the problem in the first place.’.

But the truth is, what they care more about is winning the first place in this competition. This is where the problem lies.

They are more concerned with ‘winning’ rather than whether humans can survive and society can withstand it.

AI investment of trillions, only to replace humans

In the past year, the scale of AI investment has exploded, and now the total amount has exceeded trillions of dollars.

Even Gartner predicts that global investment in the AI field will reach $2 trillion by 2026.

Regardless of who becomes the leader of the AI competition, Hinton believes that the risk of everyone being replaced has been increasing in the past year, especially ordinary workers.

An analysis shows that since OpenAI launched ChatGPT, the number of job vacancies has plummeted by about 30%, and the stock market has risen by 70%.

Last week, Amazon announced a 4% layoff, leaving 14000 employees unemployed overnight, which is the tip of the iceberg driven by AI investment.

The host asked, “So, where will such a large amount of funds flow to? Who will ultimately benefit

Hinton bluntly stated that tech giants clearly do not blindly invest and believe there will be returns.

Of course, there may also be some arrogant factors – they all want to be the first to implement ASI, even if it could kill everyone.

As for profit, besides charging chatbot fees, the main way to make money is to replace manual labor with AI.

This is the true driving force behind making the company earn the highest profits. That’s all.

Isn’t this creating a winner takes all situation?

In Hinton’s view, we are currently in an unknown territory. We have never faced an existence that is almost as intelligent as humans, or even smarter.

During the Industrial Revolution, although the steam engine was powerful, humans could always control it. In the era of AI, everything is different.

If you were replaced by robots in the past, you could still work as a customer service representative, but now even customer service representatives have been replaced by AI. Where else can these people go?

Economists optimistically believe that every technological revolution comes with the creation of new jobs.

But Hinton countered that this time is different, AI will trigger large-scale unemployment——

To make money, we need to replace human labor with AI.

Ironically, the AI dividend will be concentrated on a few individuals. Hinton made an analogy, ‘Musk will only get richer and people will lose their jobs.’.

However, this is not AI’s fault, but a flaw in social organizations.

If AI really causes massive unemployment, the only good news is that countries are united on one point – none of them want AI to take over human jobs.

Control super intelligence, humans need to be ‘babies’

Hinton believes that major companies are now using the wrong model. Executives fantasize about themselves as’ bosses’ and AI as’ super intelligent assistants’.

The boss says’ execute ‘, AI takes care of everything for him, and then the boss takes credit.

This sounds good, but when it’s smarter and stronger than us, it’s simply impossible to continue operating.

Looking around, is there a model that allows less intelligent things to control more intelligent things?

Actually, we already have such a model.

In the interview, Hinton proposed a feasible model for coexisting with superintelligence, which is that the baby controls the mother.

We have to accept that humans are babies and AI is the mother. Due to evolution, infants learn to control their mothers, and human mothers often prioritize caring for their infants.

Faced with the threat of AI to workers, the economy, and even the entire human race, people inevitably ask – does Hinton regret his past contributions?

The host asked ChatGPT how it would describe its relationship with the ‘father of AI’.

It says, ‘If I were a dense rainforest, Hinton would be the one who sowed the first seed and taught people how to water it.’. ”

However, when asked if it was “worth it”, Hinton hesitated.

If you could make a new choice, would you stop the birth of AI?

He said, ‘I’m not sure.’. Because AI, unlike nuclear weapons, can only bring bad things.

It can also do many good things, such as healthcare, education, productivity improvement, etc., but the problem lies in the optimization of social organizational structure.

AI’s ultimate foam is brewing

As trillions of capital flock towards AI, Hinton chooses to step forward and issue the harshest warning of ‘replacing humans’.

This technological feast may also breed a huge AI foam.

According to Harvard economist Jason Furman, investments in data centers and AI will contribute approximately 92% of the US GDP growth share in the first half of 2025.

If they are excluded, GDP growth is only 0.1%.

In the past two months, OpenAI has frequently signed large contracts with cloud service giants and GPU giants, and everything behind this has long been seen through.

The annual revenue is about 13 billion US dollars, and the IPO valuation is about 1 trillion US dollars. The foam created by OpenAI is larger than any previous foam in any field.

Well, today OpenAI signed a $38 billion order with cloud computing giant AWS.

Oracle, Nvidia, Amazon, Microsoft AMD、CoreWeave、 Broadcom and other major companies are deeply bound to OpenAI, and one move can have a ripple effect.

The real pivot behind all this prosperity lies in the NVIDIA GPU.

At the GTC 2025 conference, Lao Huang once again injected strong confidence into the market – Blackwell chips have been fully mass-produced, and the next generation of super Vera Rubin will be put into production in October next year.

The weight of this sentence pushed Nvidia’s market value past the $5 trillion mark in one fell swoop.

Not only that, the foam of AI has also entered the stage of “fried chicken”. After the GTC meeting, Lao Huang went viral on the internet for a meal with Samsung CEO Lee Jae yong and Hyundai CEO Jung Yi sun.

Huang is full of confidence in this. I don’t think there is an AI foam.

Last week, Meta released its financial report, investing $72 billion in AI related infrastructure, far exceeding expectations.

If we look at the entire Earth from a space perspective, data centers stand tall on a large field of farmland. Even real estate developers have bought haunted houses and built them beyond the budget.

Morgan Stanley predicts that in the next three years (2025-2028), major companies will spend $2.9 trillion on AI and data centers.

In reality, a research report from MIT in August stated that 95% of global enterprises have failed to apply GenAI.

On the one hand, players in the technology field and venture capitalists are constantly increasing their investment in AI; on the other hand, they have not yet found the “best mode” for true application.

Trillion dollar funds are still circulating in data centers.

To make a profit, as Hinton said, is to replace humans with cheap AI.

AI didn’t kill programmers, but it ruined these jobs

This trend has already begun to emerge.

Based on an in-depth analysis of nearly 180 million job postings worldwide (2023.01-2025.10), Revealera and Bloomberry’s CTO Henley Wing Chiu have found that an AI driven job market transformation is quietly taking place.

Blog Address:

https://bloomberry.com/blog/i-analyzed-180m-jobs-to-see-what-jobs-ai-is-actually-replacing-today/

Firstly, although AI has not caused a massive wave of unemployment, it has indeed dealt a heavy blow to some creative execution work.

Secondly, positions that require empathy, strategic thinking, or complex problem-solving skills, such as software engineering, creative director, and customer service, demonstrate astonishing resilience.

At the same time, ‘differentiation’ is everywhere:

Creative work is splitting into strategic positions (stable) and executive positions (declining).

The marketing field is splitting into traditional marketing positions (shrinking) and internet celebrity marketing positions (growing).

Senior leadership positions remain stable, middle-level managers are slightly worse off, and frontline employees are in the worst situation.

The complexity of the backend is still valued, while the commodification of frontend work has increased.

By 2025, the total number of recruitment positions will decrease by 8%

Let’s establish a benchmark: the number of recruitment positions in 2025 has decreased by 8% compared to the same period in 2024.

According to a recent report from the well-known recruitment website Indeed, job opportunities in the United States have decreased by 7.3% year-on-year, which is basically consistent with the data.

Now, let’s start with the positions with the largest year-on-year decline:

2. Creative execution declines, leadership positions remain firm

Among the top 10 positions with the largest decline, 3 are creative positions:

Computer Graphics Artist (-33%)

Photographer (-28%)

Author (-28%)

Computer graphic artists include technical artists, 3D artists, and visual effects artists; Writers include copywriters, copywriters, and technical document writers.

Following closely behind are journalists/reporters (-22%), also experiencing a decline in demand.

However, compared to the market benchmark of -8%, not all creative positions have been severely impacted:

Positions involving creative guidance/strategy demonstrate stronger resistance to the impact of AI.

Therefore, positions such as creative director, creative manager, and creative producer perform better than purely executive positions.

Similarly, positions involving more complex decisions and customer interactions also perform better:

Graphic designers need to spend a lot of time interpreting customer feedback and iterating.

The job of a product designer involves user research and making strategic decisions about ‘what to do’ and ‘why to do it’.

So, the core trend here is not a decrease in ‘creative job positions’, but a decrease in’ creative execution positions’, while ‘strategic creative leadership positions’ are performing well.

3. AI may be replacing medical recorders

By 2025, the demand for medical recorders will decrease by 20%.

The most likely reason may be the rise of AI document tools, which can now listen to doctor-patient conversations and automatically generate clinical records.

The job of a medical recorder is valuable, but it happens to fall under the category of structured document processing tasks that AI is increasingly adept at.

Now, let’s take a look at the positions with the largest growth:

4. Machine learning engineers are growing the fastest

From 2024 to 2025, the recruitment of machine learning engineers has surged by 40%, the largest increase among all positions.

And this is still based on a 78% growth in 2024.

Moreover, the entire AI infrastructure technology stack is thriving:

Robot Engineer:+11% (AI is moving from the digital world to the physical world)

Research/application scientists in the field of technology:+11% (companies are starting to build their own models, not just calling OpenAI APIs)

Data Center Engineer:+9% (all AI inference requires massive computing infrastructure)

Enterprises need researchers to develop models, machine learning engineers to deploy models, robot engineers to apply them to warehouses and factories, and data center engineers to power the entire operation.

5. The demand for senior leadership remains strong

Surprisingly, despite an 8% contraction in the overall job market, recruitment for senior leadership positions has hardly decreased.

Senior leadership: -1.7% (combined with directors, vice presidents, and C-level executives, outperforming the market by 6.3 percentage points)

Managerial position: -5.7% (outperforms the market by 2.3 percentage points, but still falls short of senior leadership)

Frontline employee positions: -9%

There is a 4-point gap between senior leaders and middle managers. Both are better than the market average, but the higher the position, the better the performance.

Among the top 10 fastest-growing positions, 5 are directors or above:

Data Engineering Director:+23%

Real Estate Director:+21%

Legal Director:+21%

Software Engineering Director:+14%

Vice President of Engineering:+12%

The possible reason is that companies are increasing their strategic leadership while becoming more critical of their operational management.

They hope to have more people to decide what to do, fewer people to manage how to do it, and fewer executors.

And those AI tools that threaten frontline employees are actually giving senior leaders stronger independent operational capabilities.

A product vice president who can quickly build usable prototypes with cursor or validate technical solutions with Claude naturally does not need so many subordinates to report to him.

6. Significant growth in internet celebrity marketing specialists

Overall, marketing positions are quite resilient, with most hovering around market benchmarks.

But there is one position that stands out: the recruitment of internet celebrity marketing specialists has increased by 18.3% compared to last year.

Similarly, this position grew by 10% last year and has shown a two-year growth pattern.

Behind this, there is a major trend related to AI.

With AI content flooding the Internet, traditional channels are losing their only remaining trust – more and more search results are inferior content generated by AI. Display ads have always been annoying, and now they may even be designed by AI.

But what about a skincare video from a TikTok creator of the same age? That feeling is still real and trustworthy.

Finally, let’s take a look at which positions are the most resilient:

7. Overall stability of software engineering positions

Despite the heated discussion about AI replacing software engineers, data shows the opposite——

The number of software engineer positions has not changed much since last year.

The demand for most engineering positions is either growing or hovering around market benchmarks.

And all of this happened in a year when more than a dozen AI programming assistants such as GitHub Copilot, OpenAI Codex, Claude Code, etc. took turns dominating.

The most obvious explanation is that AI tools are making engineers more efficient, rather than becoming redundant.

When you give developers a Copilot, they are not eliminated – they can deliver features faster, solve more complex problems, and spend less time on prototype code.

An interesting data point is that among all software engineering positions, front-end engineering positions have the largest decline.

This may be due to the influx of “atmosphere programming” tools such as Reply, Lovable, and Bolt.new, making it extremely easy to create front-end for websites or applications.

8. Customer service has not been replaced on a large scale

If there is a position widely believed to be eliminated by AI, it is customer service representative – demand has only decreased by 4.0%.

Indeed, chatbots can handle simple queries, but when it comes to matters that require judgment or empathy, they are completely powerless.

Especially when customers are angry or confused, they need someone who can understand their frustration, rather than a robot following a script.

High quality customer service includes empathy and occasional discretion, such as reducing fees or processing refunds.

9. Sales positions have increased or decreased

Overall, the performance of sales positions is better than the market benchmark of -8%.

Most sales positions have only slightly decreased, or even increased:

The revenue director position saw a 10.2% increase, making it the only sales position to demonstrate significant growth.

In contrast, sales and operations specialists experienced the largest decline, reaching 8.0%.

The reason may be that the latter focuses on CRM management, analysis, and process optimization; And AI has already performed quite well in handling such structured and data intensive tasks.

reference material:

https://x.com/WesRothMoney/status/1985338673894678857

https://x.com/JonhernandezIA/status/1985056068611764265

https://www.youtube.com/watch?v=e1Hf -o1SzL4

https://bloomberry.com/blog/i-analyzed-180m-jobs-to-see-what-jobs-ai-is-actually-replacing-today/