The Manhattan Project is here! 5 billion dollars to Ilya and 250 billion dollars to OpenAI

  It's the first time in the history of the world.

Something big is happening!

After two years of hiding, Ilya has finally appeared.

SSI, the security superintelligence laboratory he founded, officially announced a long-term strategic partnership with Intel.

When the news came out, the industry was shaken.

It's not just a simple shopping cart collaboration.

It is reported that INVEDA will invest "substantial" $5 billion in SSI and pledges to increase the availability of SSI by 10 times over the next 12 months.

This means that the sword of security superintelligence in Ilya's hands is finally coming out.

The fantastic ideas of genius scientists about "security superintelligence", the large-scale verification in the laboratory that can't be stopped by a calculating bottle, are finally running at full speed.

And right after that, there was another surprising news coming out.

According to WSJ, Intel is considering offering up to $250 billion in price guarantees to OpenAI to help it build its next supercomputing center with a total value of up to $500 billion and power consumption of up to 10 gigawatt.

Whether it's the Manhattan Project or the Apollo moon, it's the little witch and the big witch in front of the old yellow chessboard.

What's up with the old man?

Today, after almost two years of silence, Ilya has finally announced his return to the world.

"Our research has reached a critical point that deserves to be scaled up", he said.

At the end of the 23rd year, after a court battle, Ottoman Lightning returned and Ilya left OpenAI after six months.

The mysterious company, which hasn't published a paper or launched any products, was initially funded with a $32 billion valuation of a whopping $2 billion.

Ilya's goal is to develop a secure superintelligence.

Prior to this collaboration, SSI had been a mysterious "black box".

But from today, SSI will be different!

According to the details of the collaboration announced by the two sides, INVEDA has made a "substantial" investment in SSI amounting to $5 billion.

SSI will receive priority access to the next-generation Vera Rubin hardware platform from INVIDA.

This means that in the next 12 months, SSI's computing capabilities will make a leap in quantity!

In the interview, Ilya revealed an exciting amount of information

We've already made some breakthroughs in research, and we've found a new peak that we need to climb, which is slightly different from what I've done before.

As early as last November, Ilya suggested that the marginal benefits of LLM based solely on stacks of data and computing are declining and that humanity is "returning to a research-driven era".

Does research drive mean that you don't need mathematics?

Ilya means that we need to reopen the storm-like training on a massive scale under a completely new architecture and paradigm.

SSI believes that it has found this new paradigm (which is very likely to be different from Transformer's new architecture, or a new path to reinforcement learning and reasoning) and is "worth amplifying".

SSI may have created an early system that had the following capabilities:

• Learn new skills with little experience

• Be aware that current thinking is failing before you arrive at a final answer

• Updating itself on the premise that it is not already capable of covering

• Transfer experience from one scene to a completely different problem

• Continuing learning after deployment

Think of it as a talented teenager: it doesn't have to know all the fields at first, but it can learn all the skills quickly.

And there's hardly any evidence to support this.

Ilya has pointed out that the fundamental bottleneck of current AI is that it is too weak to generalize, the model being "overwhelmingly inferior to humans". At the same time, he described SSI's vision as creating a "super-intelligent 15-year-old" capable of learning the skills needed in any profession.

When asked how to achieve this kind of human learning ability, he admitted that he did have an opinion about some of the principles of machine learning, but declined to discuss them publicly.

The WSJ reported that the secret research direction of SSI is focusing on "the long-neglected side of how the human brain works".

Some netizens speculate on their possible technical routes:

Experience → inner judgment → self-correction → persistent learning → migration → repetition.

Inner judgment may be a brain-inspired component: humans do not wait until they have completed a 10,000-step task to know they are failing, and emotions, intuition, and judgment provide sustained feedback.

Ilya believes that AI needs an equivalent internal "value function".

Hu Jintao and Ingrid's technical team, after seeing the internal research at SSI, immediately decided to invest heavily.

"Ilya has already made pioneering breakthroughs in the foundations of modern AI, and we look forward to seeing what new breakthroughs SSI will discover with our Vera Rubin platform!"

Today, Ilya's lab officially received the world's top-notch unlimited firepower experience card.

With a market cap of $4.7 trillion, why go out there and spend $5 billion investing in an SSI that doesn't even have a product?

A well-known Silicon Valley investor, Dan McAteer, gave an elaborate analysis of this.

SSI stands for NVIDIA and SSI stands for NVIDIA.
It's just like DeepMind on Google.
OpenAI is for Microsoft, and OpenAI is for Microsoft.
It's the first time I've ever heard of anthropic.
Cursor is owned by SpaceXAI.
Every giant in mass computing must have its own cutting edge laboratory.

This argument exposes the underlying codes of the capital game.

Measuring the smart age will be a simple core indicator:How much intelligence can unitary computing produce?Whoever wins this will win the future.

For more than a year, Huang Yin has been "sweeping" the craze. He not only invested in SSI but also bet big money on Thinking Machines Lab, the former CTO of OpenAI.

It's clear that the Great Wall of England can't just be built on hardware.

Previously, AMD had just announced an investment of up to $5 billion in Anthropic.

If all the top AI labs were divided by Microsoft, Google, DeepMind, Amazon, and Anthropic, then the future of Great Britain would be a purely hardware substituted factory that could be popped out at any time by these giants of self-developed chips like Google TPU and Microsoft Maia.

Therefore, Ingrid must deeply tie the brains of the world's smartest.

Investing in SSI is the ultimate ferry ticket for INVIDA to ASI.

In addition, the company has been exposed to more than this.

Recently, according to media outlets such as WSJ, India is in secret negotiations with OpenAI, with plans to provide OpenAI with up to Rs.$25 billionOpenAI, the world's most powerful supercalculation center, is being used to "pack up" OpenAI!

How big is this project?

Total cost:More than $500 billion (about 36 trillion yuan)

Electricity needs:10 gigawatt

Developers:Energy subsidiaries of SoftBank Group

Location:Southern Ohio, USA

This astonishing project challenges the limits of human engineering history.

10 gigawatts of electricity is enough to meet the full electricity needs of a 10 million-strong supercity (like New York City).

This may require the construction of a number of special nuclear power stations and government approval.

What is the concept of $50 billion?

The Manhattan Project is estimated to cost about $30 billion; the Apollo Moon Project is estimated to cost about $250 billion.

And this supercomputing center, built by OpenAI and SoftBank, will cost tens of times the cost of the Manhattan Project!

It is interesting to note that the role played by Ingrid is very subtle.

The $250 billion is not a direct backpack for Ingrid, but a "guarantee", backing up its own balance sheets and credit ratings.

If SoftBank goes to the bank and the banks fear that OpenAI won't pay the rent, the UK could collapse.

In addition, Intel is reportedly also in separate negotiations to provide $350 billion in financing support for the acquisition of OpenAI's chips.

It's a perfect game of calculation.

1. UKDA guarantees OpenAI that SoftBank will build a data center.

2. OpenAI borrowed the chips to fill this data center.

Who bought the chips? Of course it was Ingrid!

The $400 billion went around the corner, eventually turning into a blinding revenue and profit on the profits of British Airways.

So, not only did Great Britain lock up OpenAI, the world's biggest "Golden Beast", but it also tied SoftBank, Wall Street capital, to its own tanks.

If you look at the $5 billion guaranteed by SSI and the $250 billion guaranteed by OpenAI, you'll find that Old Yellow is too forward-thinking.

The current development of AI is at the crossroads of two routes:

Route one: Aesthetics of Violence, The Force Awakens on behalf of OpenAI.

In this way, as long as the computing power, the stacks of data, can emerge from the AGI, what is needed is a building frenzy.

Route two: paradigm revolution, security superintelligence (represented by SSI)

The marginal effect of scaling has been reached, and new model architectures and breakthroughs are needed.

What did the old man choose?

It's the little boy who makes the choice, and he's the one who makes the choice.

OpenAI, with a leverage of $50 billion, has completed an unprecedented build-up, so GPU sales will remain astronomical for years to come.

For SSI, Old Yellow used Vera Rubin to buy a "technical paradigm shift" super-insurance. Even if the power grid flew to the brink, Ilya's new breakthrough could keep Britain as a growth engine for the next decade.

INVIDA has gone from being a simple seller to becoming the central bank of the AI era.

Whether you are Ilya or Ottoman, on the ASI card table, you have to use the chips issued by Huang Ren.