You provide the capital and the data, we'll co-own the data centers share the models until Trump and the US government decide to shut it off as a bargaining chip.
The article reads "some four billion human beings alive today—can trace their ancestry to the Yamnaya". It's 50%. I haven't checked the research though.
> r1b's population is only 190 million [1]
Their population likely had multiple haplogroups. R1b was the most common (5000 years ago). But, yeah, the 50% number looks too high.
> One commonly used (though also criticized) benchmark for housing affordability is that no more than 30% of household income should go toward housing costs. Households that spend more than that are considered “cost burdened” by the U.S. Department of Housing and Urban Development.
> By that standard, 31.3% of American households were cost burdened in 2023, including 27.1% of households with a mortgage and 49.7% of households that rent, according to 1-year estimates from the Census Bureau’s American Community Survey (ACS). (Many more people own than rent: In the second quarter of 2024, 65.6% of occupied housing units were owned while 34.4% were rented, according to the most recent estimates from the Census Bureau’s Current Population Survey/Housing Vacancy Survey.)
Only looking at home prices compared to salary is very misleading because it doesn't account for changes in interest rates. Mortgages were almost 20% interest in the 80s. Cheaper doesn't mean much if you still can't afford the monthly payment.
Also looking at average price doesn't account for the rising quality of housing. In the 1980s the average home was around 1,700 square feet. Today, it is nearly 2,700.
If you look at the pricing trend of a single house, it tells quite a story
In my city, A house that would have been 80k in the 80s is listed between 500-600k today, depending on the neighborhood and how updated it is
In the 80s you could get a 15-20 year mortgage at 20%
Now you get a 30 year mortgage at 5%
If your monthly payment today is less than it would be at 20%, it is only because you are expected to be paying for it at least an extra 10 years compared to the past
There is absolutely no question that houses are less affordable today than they used to be
And that's before even thinking about how salaries haven't grown anywhere near as quickly as real estate prices
An 80k house in the 80’s means inflation alone accounts for ~$300k of the current sale price of the house. If the general area has built up at all in the last 40 years, that could account for a bunch more of that cost. Absolutely some areas and places are climbing way faster than their market wages are keeping up, but I also think a lot of housing discussion compares a house 1 hour outside of the nearest big city with that same house now in the middle of that expanded big city. Location matters a lot, and what is a great location now might well have been out in the sticks 40 years ago.
> In the 80s you could get a 15-20 year mortgage at 20%
20% was the rate for 30 year mortgage in the 1980s. My source is specifically for 30 year mortgages.
> If your monthly payment today is less than it would be at 20%, it is only because you are expected to be paying for it at least an extra 10 years compared to the past
That's a gross overgeneralization. Interest rates are lower across the board today.
> If your monthly payment today is less than it would be at 20%, it is only because you are expected to be paying for it at least an extra 10 years compared to the past
I never said they weren't but you also haven't provided any evidence that arent.
> And that's before even thinking about how salaries haven't grown anywhere near as quickly as real estate prices
You're literally just repeating your original claim with no new evidence.
It really isn't relevant in the way that you think.
The fiscal cycle is a ponzi cycle following a ponzi curve this mirrors many areas including business growth S-curves.
In general that characteristically means benefits start front-loaded, they have a period of diminishment, and then at a point outflows exceed inflows where you have to pay back and keep paying more than you spent. The overall plan being by the third stage, you are dead and don't have to pay it back, or have been bought out and its someone else's problem.
Even with normalization of the price level it doesn't accurately reflect purchasing power well, and so you cannot really measure opportunity cost or make an accurate objective comparison.
This is the nature of fiat money distortions and why Mises was so against Socialism. In his works he defines the Economic Calculation Problem, or the Socialist Calculation Problem whichever you rather prefer. Sustaining chaotic distortions are ECP/SCP.
The nature of fiat/ponzi is that money printing in an economy debases exchange, extracting cost and forcing failures broadly to non-market socialism when that third stage happens.
This is reflected in a lack of employment, and no or gradually fewer businesses entering the market/industry. It acts like a sieve, with the money printer continuing even after no profit can be made (where regular business exits the market). It is a parasite that kills its host every time, but that is a problem for next quarter.
Distortions take many forms and they are chaotic and by that nature they unknowable in detail specifically and unpredictable. Artificial Supply Constraint to raise price level is one such form.
Often there are whipsaw dynamics between opposing constraints, with diminishing returns required to remain stable. A cliffside with drop-offs on either side as you approach, and you can only march forward into the ocean. At first there is a minor safe path, but eventually it converges. Hysteresis can be an impossible to solve problem without being able to change the underlying system.
Price levels being suppressed (such as Gold/Silver/Food). Price discovery being manipulated (dark pool transactions exceeding exchange volume). These are signs of chaotic distortions caused by money printing.
Economic calculation requires price discovery, which requires adversarial decision-making. Cooperative behaviors naturally occur when there are few participants. The distortions injected through the money supply have knowable (in general) dynamics and outcomes.
Its important to remember that Price != Purchasing Power. Wage suppression is also a real thing and its fueled by the same.
The authors make a case that China's GDP is overestimated from past night time studies so I ran an analysis on which country produces the most nighttime lights.
China’s population is packed into dense urban areas and high-rise buildings, unlike the sprawling suburbs of the US... Could this mean fewer distinct “dots of light” on the map, even if economic activity is high? It looks like measuring overall light intensity is a part of the analysis as well which should help counter this...
They baited me into putting in a query and then asking me to sign up to submit it. Even have a "Stay Logged Out" button that I thought would bypass it, but no.
I get running these models is not cheap, but they just lost a potential customer / user.
super impressive. we won't need that many GPUs in the future if we can have the performance of DeepSeek R1 with even less parameters. NVIDIA is in trouble. We are moving towards a world of very cheap compute: https://medium.com/thoughts-on-machine-learning/a-future-of-...
Have you heard of Jevons paradox? That says that whenever new tech is used to make something more efficient the tech is just upscaled to make the product quality higher. Same here. Deepseek has some algoritmic improvements that reduces resources for the same output quality. But increasig resources (which are available) will increase the quality. There will be always need for more compute. Nvidia is not in trouble. They have a monopoly on high performing ai chips for which demand will at least rise by a factor of 1000 upcoming years (my personal opinion)
I agree that the Jevons paradox can apply here, however, there have been several "breakthroughs" in the last couple of months (R1, diffusion LLMs, this) that really push the amount of GPU compute down such that I think it's going to be problematic for companies that went out and bought boatloads of GPUs (like OpenAI, for example). So while it might not be bad news for NVidia (given Jevons) it does seem to be bad news for OpenAI.
Even if you have cheaper models if you have tons of compute power you can do more things than if you had less compute power!
You can experiment with huge societies of agents, each exploring multitude of options. You can run world models where agents can run though experiments and you can feed all this back to a single "spokesperson" and you'll have an increase in intelligence or at the very least you'll able to distill the next generation models with that and rinse and repeat.
I mean I welcome the democratizing effect of this but I fail to understand how this is something that is so readily accepted as a doom scenario for people owning or building massive compute.
If anything, what we're witnessing is the recognition that useful stuff can be achieved by multiplying matrices!
> I guess the investors selling NVIDIA's stock like crazy know nothing about jevons
I know you are trying to be sarcastic, but for the sake of argument let's assume that your question is genuine.
There are two types of investors and they both sell, but for different reasons:
1. Casual investors: They don't know much about investing, or Jevons paradox. They only watch the news, so they panic sell.
2. Institutional investors: They know all about Jevons paradox etc, but they also know that casual investors don't, so they sell on purpose so that they can buy the dip later.
Surprisingly those open models might be savour for Apple and gift for Qualcomm too. They can finetune them to their liking and catch up to competition and also sell more of their devices in the future. Longterm even better models for Vision will have problem to compete with latency of smaller models that are good enough but have very low latency. This will be important in robotics - reason Figure AI dumped OpenAI and started using their own AI models based on Open Source (founder mentioned recently in one interview).
You provide the capital and the data, we'll co-own the data centers share the models until Trump and the US government decide to shut it off as a bargaining chip.