AI will strain the power grid to the breaking point. NetZero will completely destroy the grid. These two mass movements will try to rip each other apart. The only winner in the end is likely to be the billionaire elites with the most cash and energy reserves, and real world soldiers. Because after these brats do their worst, a real war is not out of the question.
Artificial intelligence is beginning to emerge from the primordial ooze in which it has been evolving over the past 75 years. It is growing some appendages, and perhaps a prehensile thumb or two. But it’s not yet ready for prime time, and it ain’t gonna come cheap, that’s for sure! The trend is toward $trillion dollar compute cluster installations, which may require as much electric power as 20% of US national demand! Power requirements will prove a significant constraint.
The term “OOM” below refers to “orders of magnitude” improvement in machine cognitive power.
The image above is taken from the Leopold Aschenbrenner essay “Situational Awareness,” where he warns about the rapid avalanche of cognitive capacity for the near future of machine intelligences. More
The Aschenbrenner essay has been explained on multiple websites and on several YouTube videos. The Matthew Berman video below is one that is fairly well presented:
Here is a summary list of some of the key points alluded to above:
Key Points at a Glance
- Super intelligence surpassing human intellect is imminent.
- Rapid advancements in AI technology are driving towards AGI.
- Transformers’ architecture may limit scaling to AGI.
- Skepticism around AI capabilities continues to be disproven.
- Compute advancements drive significant AI progress.
- Algorithmic efficiencies play a crucial role in AI evolution.
- Data scarcity poses a challenge for AI development.
- Transition from chatbots to intelligent agents is imminent.
- Importance of allowing models time to process prompts.
- Potential of AI models to autonomously improve through iteration.
- Implications of superintelligent AI on military and economic domains.
- Challenges and risks associated with superintelligent AI.
- Risks of inadequate security measures in AI development.
- Securing AI labs is crucial for national security.
- Importance of safeguarding AI advancements against foreign exploitation.
- Challenges in aligning super intelligent AI with human incentives.
- Super alignment approach involves cascading model supervision.
- Government involvement in AGI development is deemed essential.
- AGI development requires a shift akin to nuclear arms development.
Each point in the list is explained further here.
The current batch of AIs are very fallible. They have an unfortunate tendency to hallucinate and tell outright lies and fabrications.
But it would be a mistake to judge next year’s AI by last year’s AI mistakes and shortcomings. Every year is seeing greater functionality for the many new generations and varieties of AI creations.
Some Investors Are Making Big Plans for AI
From a money-making investment point of view, AI can make most aspects of doing business much more efficient. And that will save $billions a year for large companies. That kind of savings in operations tends to attract investment. Nvidia has already become the most highly valued corporation, beating both Microsoft and Apple.
Estimates vary on how much investment AI will drive, but most believe it will result in additional trillions of dollars in investment over the next decade, UBS said. As such, the implications for companies, workers and investors are enormous.
“While it is too early to accurately quantify the aggregate productivity enhancements from AI, anecdotal evidence suggests substantial efficiency gains,” UBS wrote.
Analysts at the firm laid out a series of predictions for how AI will reshape the global economy and their broader implications.
1. AI will be the most profound innovation and one of the largest investment opportunities in human history.
2. AI will kick off a data center capex cycle that will dwarf general purpose data center capex in the next years.
3. The ratio of monetization of the AI application layer to the costs of the enabling and intelligence layers will become a key metric for investment returns.
4. The race to artificial general intelligence could trigger a capex cycle that inflates an investment bubble where the capex of the enabling layer is dissociated from near-term monetization potential of the application layer.
5. The AI enablers will be the first adopters of AI, driving both revenue and margin upside.
6. Monolithic players will emerge along the AI value chain and over time, the AI market will be dominated by an oligopoly of vertically integrated “AI foundries.”
7. The AI silicon moment: AI chips will capture a large part of the AI value creation.
8. The application and intelligence layers will merge with AGI.
9. Software will become ubiquitous.
10. Data assets will emerge as the competitive differentiators for AI adopters information technology.
10 Predictions How AI Will Reshape the World (from an investment POV)
Four and a Half Hours of Leopold Aschenbrenner!
In the following 4 1/2 hour video interview, Leopold Aschenbrenner elaborates on the many details hidden within his provocative essay.
The idea that AI threatens to shake the modern foundations of our world, is not a new idea. Stephen Hawking, Elon Musk, Ray Kurzweil, and many other famous visionaries have been repeating the warning (and the promise) for decades. But not many people actually believed them.
For many years, I have believed that a machine actually had to be intelligent before it earned the name “artificial intelligence.” But of course, that is not true. Garry Kasparov learned that very lesson, when he was defeated at the chess board by a not-very-intelligent supercomputer. To be called an AI, a machine only needs to perform a narrow task at human-level or better.
What is beginning to concern a larger group of persons today is the possibility that machines might learn to become conniving self-centered thugs, with the intelligence of ultra-bright super villains. But if we were truly worried about that possibility, why would we build $trillion dollar computing clusters for the things?
Think about it. The reason is the same one that drives wealthy visionaries to invest in expensive space launch and outer space infrastructures and systems. Because it is a way for a $billionaire to become a $trillionaire. If you read the essays linked above and watch the videos, you may see how the build-up of the fleet of artificial super-intelligences will inject huge amounts of capital into the commercial and industrial systems. The resulting economic boom could exceed what happened in the 1980s and 1990s combined.
There are still many naysayers, of course. Physicist Sabine Hossenfelder goes on the attack against Leopold Aschenbrenner on the basis of energy demands and material constraints — to say nothing of the constraints placed by climate hysteria! Remember, if a person is in thrall to the Climate Apocalypse Cult then all one has to do to win an argument is to scream out: C L I M A T E !!!
NetZero is jealous of all the $trillions in investment that AI wants to take for itself. On top of that, NetZero wants to use all the mineral output, investment capital, and energy output of the planet to displace all fossil fuel and nuclear power generation and replace it with wind and solar. The sheer suicidal idiocy of NetZero cannot be overstated.
Is AI likely to lead to human extinction? Some AIs seem to think so. But unless some malevolent elite-nihilist billionaires gain control of the AIs — and order them to cause the extinction of large numbers of humans — such a scenario is not likely.
In some ways, humans are like rats and cockroaches. They find ways to survive. And if by chance the NetZero vs. AI wars (plus all the ancillary real wars) happen to destroy most universities, media outlets, government bureaucracies, non-profit foundations, and other DIE-woke cultural institutions, the human race may just be given a new lease on life!
al fin next level 