mvanaltvorst

Is it just me, or is it really obvious that Wout won?

https://i.imgur.com/g78YNla.png

Each pixel column is a specific point in time. I added a few landmarks: when the front of the wheel passes the camera view, when the axis of the wheel passes the camera view, and when the end of the wheel passes the camera view. Blue for the top biker (Wout) and green for the bottom biker (Tom).

Every single event happened earlier for Wout. Both wheels are the same diameter. Therefore, the entire period where the wheel intersected the camera view, Wout's wheel was ahead. Including the exact period where Wout passed the actual finish line.

If Tom would have passed Wout, you would expect that the bottom wheel would be more "squeezed" such that the end of Tom's wheel would have passed the camera line earlier than Wout. Same for the other landmark (centre of wheel).

You can add more landmarks if you want (e.g. approximate when the wheel passes for 25% and for 75%), but it should be clear by pure margin that Wout had the lead for the whole duration where the front wheels passed the camera view. Pretty much all the computation in this article was unnecessary.

Amstel gold made the good call, though I agree that it was probably more luck.

The top-of-the-line Apple Watch is $529. You can go higher than that, but that is purely for aesthetics (upgrade to titanium shell, stainless steel clasp, etc.).

There's an image on their website[1], pretty huge water pumps.

"To solve the 70-year-old problem of wafer-scale, we needed not only to yield a big chip, but to invent new mechanisms for powering, packaging, and cooling it.

The traditional method of powering a chip from its edges creates too much dissipation at a large chip’s center. To prevent this, CS-2’s innovative design delivers power perpendicularly to each core.

To uniformly cool the entire wafer, pumps inside CS-2 move water across the back of the WSE-2, then into a heat exchanger where the internal water is cooled by either cold datacenter water or air."

[1]: https://cerebras.net/product/

The fastest available ARM device in the consumer Windows ecosystem, the Microsoft Surface Pro X, is significantly slower than a virtualised instance of Windows on a MacBook Air. There is a penalty, but it doesn't matter that much in practice (just reduced battery life).

You could also remove the whole AirTag?

The fact that it works with any iPhone /is/ the innovation. I would imagine less than 0,5% of iPhone users has the Tile app installed. A worldwide network of connected iPhones will make this technology finally worth using.

If you take this CPU with its 280W TDP and scale it down to 18W like an Apple M1, you get a score of ~7200 while the Apple M1 gets 7700. Actually pretty amazing how close the performance per watt is to be honest; totally different form factors, different CPU architecture, different R&D (both having spent billions of dollars), different fabs, but still within 10% performance/watt. Makes you wonder how close we are to the physical limits.

Not screwing over ~9M wallstreetbets users, possibly resulting in a short squeeze, nets you millions of loyal customers for life. On top of that, the whole reason the stock is currently priced so high is the lack of available shares. Release slightly too much new shares and stock prices will plummet. It's not a linear relationship.

China has always been extremely good for its top students, providing them with all they need to excel. Once China becomes a global superpower, I'd imagine many of them go back to working for China if they get the chance to with their new-found expertise from abroad.

>Chinese middle class will demand the same standard of living like people in the West

Will they? As far as I know censorship has been amazingly successful

That sounds interesting, but I would imagine it would become very complicated once you start applying nontrivial functions (discontinuous functions, for example). In that case the range of possible values could actually become discontinuous. I would imagine accounting for that is actually more computationally expensive than just using arbitrary precision decimals.

Correct, and judging from someone else in this thread, it was even possible to use wildcard matching to get access to an entire list of users at once.

People have gotten taller and healthier because of better nutrition, there is no reason to believe people haven't gotten smarter as well.

Do you really feel like you would be constrained by the amount of shortcuts on a keyboard without a numpad? There are so many different combinations you can make once you include control/alt/fn etc. Sure, if you're used to shortcuts on a numpad I can see how it might hinder your productivity for a few weeks. However, working on a smaller keyboard will definitely not constrain the amount of flexibility you have in the long term once you get used to other shortcuts. You might even become more productive on a smaller keyboard, because it takes less effort to stay on the home row and find keys.

You don't have to know how much gas it requires to /complete/, you can just run the program and check whether it runs out of gas before it exits. You do not need to solve the halting problem. Actually, it's pretty trivial to implement this and this model has scaled quite well considering the size of Ethereum today.

I have never had carpal tunnel-like symptoms yet, though I just realised I've only used left ctrl for the past 10 years. I'll definitely be more mindful of that. Do you have any more resources on avoiding common typing pitfalls, or specific wrist movements one should avoid while typing? Should capitalising your sentences be a two-hand action as well, considering I normally only use left shift?

What would the implications of FHE be for the average consumer?

This effectiveness of product has been debunked half a year ago [1], though the technology is very interesting.

[1]: Bui, S., Geitung, L., Oppedal, F., & Barrett, L. T. (2020). Salmon lice survive the straight shooter: A commercial scale sea cage trial of laser delousing. Preventive Veterinary Medicine, 181, 105063. doi:https://doi.org/10.1016/j.prevetmed.2020.105063

Take the money, pile it all up and directly give it back to the citizens, where every citizen gets the same amount of money. Effect: households will have the same amount of money on average, but people will be nudged to substitute their carbon-intensive lifestyle by greener alternatives. (or, in economic terms, there is no income effect, only a substitution effect).

You're also indirectly taking money from big corporations and putting it into the pockets of citizens, which has been proven to be one of the most effective methods of stimulating the economy.

> solve [...], otherwise everything will keep going up

This implies that markets going up is a problem that needs to be solved.

TSMC

Apple could use DDG as leverage though. If Google stops paying those 9bn, Apple has no reasonable alternative search engine it could resort back to. By quietly continuing development on DuckDuckGo for ~500mn/year, it could be some deterrence towards Google's monopoly that aligns with Apple's own interests.

Huh, you're right! The current #500 system [1] consists of 1496 20-core Xeon processors, for a linpack performance of 1316TFLOPS. The GPU of an iPhone 12 is ~800GFLOPS for FP32 on the GPU, which implies that ~1600 iPhone 12's combined are more powerful than this cluster of Xeon's when you look at raw compute power.

The difference is that a Xeon is significantly more flexible than a GPU, though. I cannot find good performance numbers on the A14's CPU, though an estimated guess suggests that it would be more in the order of magnitude of ~10GFLOPS instead of ~800GFLOPS per A14. Or ~100.000 iPhones to match the #500 supercomputer.

Apple sold ~40 million iPhone 11's in 2020, with iPhone 12 sales being ~20% higher than those of the iPhone 11. That's about 10 supercomputers worth of iPhone CPU power sold per week.

[1] https://www.top500.org/system/179351/

You differentiate both sides of the equation given above with respect to t. Differentiating 1/t yields -1/t^2. On the LHS you apply the differentiation inside the integral.

I prefer this one: http://www.math.wm.edu/~leemis/chart/UDR/UDR.html

Of course, I had accumulated all other pruning strategies under the "other optimisations".

You can imagine playing chess as a tree. You start with a game state, and every possible move is an edge to a new game state. Unfortunately, the amount of game states grows approximately exponentially (e.g. for every game state there are approximately 40 moves to play, thus every extra move multiplies #(game states) by 40, approximately). This neural engine is a trick so that Stockfish does not have to simulate all 40 moves every game states. The neural net outputs the moves that are most likely to be strong moves, and Stockfish will only consider those moves. Of course, this is a very basic explanation and there are many more optimisations Stockfish uses, though the ultimate goal of almost every optimisation is to reduce the amount of simulation Stockfish has to do.

Say that Pfizer were able to simulate 50 possible mutations and put all of them into a single vaccine, would there be any risk that the antibodies could interact with each other in some way that makes the vaccines less effective?

Also another scenario: say that we get yearly Covid shots that contain the latest strains, would these accumulate over the years and start interacting with each other?

Gradient descent is a subset of survival of the fittest, described by Darwin in 1800-1900, and has been in applied in computer science since the 70's. An AGI will probably use some form of gradient descent during its training, yes, but I wouldn't argue that this has brought us even close to an AGI.

I'm curious, which new competitors do you mean precisely? Do they have better capabilities than Capella? This seems like a great upcoming industry