Most OLEDs today ship with logo detection and will dampen the brightness on static elements automatically.
While it isn’t a silver bullet, it does help reduce burn in since it is strongly linked to heat, and therefore to the pixel brightness. New blue PHOLEDs are expected to also cut burn in risk. Remember that LCDs also used to have burn in issues, as did CRTs.
I’ve been using Nvidia under Linux for the last 3 years and it has been massive pita.
Getting CUDA to work consistently is a feat, and one that must be repeated for most driver updates.
Wayland support is still shoddy.
Hardware acceleration on the web (at least with Firefox) is very inconsistent.
It is very much a second-class experience compared to Windows, and it shouldn’t be.
Linux and Nvidia really need to sort out their shit so I can fully dump windows.
Luckily the AI hype is good for something in this regard, since running gpus on Linux servers is suddenly much more important.
Humans are intelligent animals, but humans are not only intelligent animals. We do not make decisions and choose which beliefs to hold based solely on sober analysis of facts.
That doesn’t change the general point that a model given the vast corpus of human knowledge will prefer the most oft-repeated bits to the true bits, whereas we humans have muddled our way through to some modicum of understanding of the world around us by not doing that.
But the most current information does not mean it is the most correct information.
I could publish 100 papers on Arxiv claiming the Earth is, in fact, a cube - but that doesn’t make it true even though it is more recent than the sphere claims.
Some mechanism must decide what is true and send that information to train the model - that act of deciding is where the actual intelligence in this process lives. Today that decision is made by humans, they curate the datasets used to train the model.
There’s no intelligence in these current models.
Copilot is GPT under the hood, it just starts with a search step that finds (hopefully) relevant content and then passes that to GPT for summarization.
With refresh rates like that, you must be talking about LED billboards.
These are different from consumer monitors, which mostly use constant LED backlights and a liquid crystal layer to determine color.
An LED bilboard is going to have a fuckton of singular LEDs - each of which can emit exactly one color - arranged in groups to form full pixels capable of displaying many colors. There is no extra LCD layer between your eyes and the billboard LEDs.
The reason for the high refresh rates is because each led must be extinguished and and relit to redraw the image, and the eye is very good at picking up this strobe effect.
The difference vs. a consumer display is that the backlight in a typical monitor is constant. Refreshes the screen involves sending updated instructions to the LCD layer, twisting the crystals and possibly changing the color they allow through.
To make a crude concrete example:
Imagine I am shining a white flashlight in your face. In front of the flashlight I put a colored piece of plastic so the light hitting you is colored. Then I change the plastic to one with a (slightly) different color. I do this 120 times per second. That is a typical consumer display.
Now imagine I am shining a colored flashlight directly in your face. Then I turn it off and grab a flashlight of a different color and shine it in your face. Imagine I do that 120 times per second. That is an LED billboard.
Which do you think is more likely to give you a headache?
One final complication - the brightness of the LEDs is variable over time, they received a modulated signal rather than a steady voltage, so at lower refresh rates there will be a noticeable ripple across the image, similar to how early CRT screens could look.
Increasing the refresh rate hides a lot of these problems.
Every billion parameters needs about 2 GB of VRAM - if using bfloat16 representation. 16 bits per parameter, 8 bits per byte -> 2 bytes per parameter.
1 billion parameters ~ 2 Billion bytes ~ 2 GB.
From the name, this model has 72 Billion parameters, so ~144 GB of VRAM
Check it out to throw in the trash. Jared Diamond’s book is thoroughly condemned in anthropological and archaeological circles.
it’s not spontaneous
Spontaneity in thermodynamics refers to a process which occurs without external application of energy. In your description, a pile of ash becoming an apple is spontaneous.
So in a contained universe, it doesn’t matter if it’s an apple releasing energy and becoming a pile of ash, or a pile of ash absorbing energy and becoming a perfectly normal apple.
The net energy is still conserved. Just going from energy to mass unlike mass to energy.
There is no mass-energy conversion in an apple burning to become ash, just the release of chemical energy from newly-formed bonds.
Regardless, conservation of energy is only one part of how the universe operates. The second operating principle is (or at least from hundreds of years of scientific inquiry appears to be) the maximization of entropy. That is the ‘spreading out’ of available energy. This is the reason iron rusts, rather than remaining oxygen and iron - conservation of energy alone cannot explain natural phenomena.
Spontaneous reconstruction of an ashed apple violates the second law of thermodynamics, and the Second law is no less valid than the First.
Lastly, I was not writing specifically about Penrose’s views on consciousness. His entire theory that gravity is driving the collapse of a wave function, and that said collapse occurs retroactively, is untested and based on an appeal to elegance. This does not make it wrong, but it most certainly should not be taken as true.
Beyond consciousness, the second law of thermodynamics also implies the presence and direction of time. In fact, it is sometimes called the Arrow of Time as it appears to direct physical processes to happen preferentially in the direction that increases entropy.
A self contained universe with fixed energy and infite time will eventually see a pile of ash turned into an apple. And it wouldn’t violate a damn thing with our system of physics.
This occuring spontaneously would indeed violate the 2nd law. This is a core disagreement between classical thermodynamics and statistical mechanics, which seems to re-derive classical thermo from probabilistic arguments over system states.
I feel it also warrants stating that Penrose’s theory is not widely accepted, has yet to be tested, and is based mostly on an argument to elegance - it “seems weird” for their to be uncountably infinite parallel timelines spawning at every instant. It is far too soon for it to be taken as fact.
Explaining what happens in a neural net is trivial. All they do is approximate (generally) nonlinear functions with a long series of multiplications and some rectification operations.
That isn’t the hard part, you can track all of the math at each step.
The hard part is stating a simple explanation for the semantic meaning of each operation.
When a human solves a problem, we like to think that it occurs in discrete steps with simple goals: “First I will draw a diagram and put in the known information, then I will write the governing equations, then simplify them for the physics of the problem”, and so on.
Neural nets don’t appear to solve problems that way, each atomic operation does not have that semantic meaning. That is the root of all the reporting about how they are such ‘black boxes’ and researchers ‘don’t understand’ how they work.
Those without respiratory issues don’t have oxygenation issues while wearing masks.
The same air is going into your lungs, the only difference is your diaphragm has to work harder due to the filtering effect of the mask. If it can’t manage that, then you are likely already on oxygen due to low tidal volume and chronic hypoxia.
You can buy a pulse oximeter from CVS for like $20 and test this yourself if you don’t believe me.
In the language of classical probability theory: the models learn the probability distribution of words in language from their training data, and then approximate this distribution using their parameters and network structure.
When given a prompt, they then calculate the conditional probabilities of the next word, given the words they have already seen, and sample from that space.
It is a rather simple idea, all of the complexity comes from trying to give the high-dimensional vector operations (that it is doing to calculate conditional probabilities) a human meaning.
No, it isn’t. The key conceit is they are removing water from the river and evaporating it.
The water isn’t ‘lost’ it is still part of the hydrosphere, but it is made non-local. That water goes into the air and will go on to be rain in some place far away from the community where it was sourced. This will absolutely contrubute to local droughts and water insecurity.
There are really only 3 search providers, Google, Bing, and Yandex.
All others will pay one of these three to use their indexes, since creating and maintaining that index is incredibly expensive.