Glitches may appear if simulators use very simple world-modelling systems, like 2D surface modelling instead of 3D space modelling, or simple neural nets to generate realistic images like our GANs.
avturchin
Karma: 1,842
I now have covid after being vaccinated 3 months ago by Russian Sputnik-V vaccine. For now, it is mild: one day of 38 C, 3 days of 37.5C, only upper level infection, no cough. I lost smell, but it is is slowly returning. Oxygen at my normal level.
Yes, it is a reversed doomsday argument: it is unlikely that the end is nigh.
I agree that simpler simulations are more probable. As a result the cheapest and one-observer-cenetered simulation are the most numerous. But cheapest simulation will have the highest probability of glitches. Thus the main observable property of living in simulation is higher probability to observer miracles.
Wrote about it here: “Simulation Typology and Termination Risks” and Glitch in the Matrix: Urban Legend or Evidence of the Simulation?
Everything exists in the multiverse, so there is no choice: a child will be born anyway
Here you have neurons and order of their connection. This order is a graph and could be described as one long number. Mind states appear as a brain moves from one state to another, and here it will be transition from one number to another.
This is exactly what Muller wrote in his article, which you linked: you need just numbers and a law based on Kolmogorov’s complexity which connects them - to create an illusion of stream of consciousness. Neurons are not needed at all.
One don’t need to limit breeding ability to get genetic selection: just move those who you not interesting on into outside world. For example, i want to breed dogs who are good to live in mountains. I put a group of dogs on a mountain. As generations pass, those dogs which are good in living in the mountain will remain there and even move higher, and those who are not good in it, will move to live in plains. Those who live in plains may even have higher number of children.
If we are inside the colonisation volume, its change will be isotropic and will not look strange for us. For example, if aliens completely eliminated stars of X class as they are the best source of energy, we will not observe it, as there will be no X stars in any direction.
Grabby aliens and Zoo hypothesis
Both answers are correct: manufacturing is slow and a lot of people are against vaccination. More about in this article in Russian.
Population trust is low and many people deny vaccination. Only around 5 per cent of people in Moscow has been vaccinated, despite easy availability of vaccine for everybody in Moscow. Also, there is a typical situation in Russia from Soviet times when Moscow is oversupplied and regions are undersupplied. Ural region has run out of vaccine.In my circle, 80 per cent of people had covid, but only a few friends vaccinated (around 5-10 per cent based on the share of people who visited my last year wedding and later vaccinated). There are two explanation according to them: either they already had covid and has antibodies and don’t see the reason to take risks of vaccination, or “they are too old” and afraid of side effects. Interestingly, Russian vaccine was initially approved only for people below 60, Also, unfortunately, Putin didn’t take the shot and this didn’t help the belief in vaccine.
Two more Russian vaccines are in the late stages of approval, and some people wait for them as they are based on more conventional technology: not viral vectors, but dead coronavirus.
Russian x-risks newsletter winter 2020-2021: free vaccines for foreigners, bird flu outbreak, one more nuclear near-miss in the past and one now, new AGI institute.
as a weak alignment techniques we might use to bootstrap strong alignment.
Yes, it also reminded me Christiano approach of amplification and distillation.
Interesting hyperbolic model here: https://www.researchgate.net/publication/325664983_The_21_st_Century_Singularity_and_its_Big_History_Implications_A_re-analysis
Yes. I am working on an article about this idea. This is is especially true in the case of climate change. Runaway global warming is much more probable because of survivorship bias.
The idea of the Doomsday weapon as it was envision by Kahn is that it will be activated automatically and can’t be turned off by survivors—and it is well known fact for all players.
gwern on reddit:
“The most immediate implication of this would be that you now have a scaling law for transfer learning, and so you can predict how large a general-purpose model you need in order to obtain the necessary performance on a given low-n dataset. So if you have some economic use-cases in mind where you only have, say, _n_=1000 datapoints, you can use this to estimate what scale model is necessary to make that viable. My first impression is that this power law looks quite favorable*, and so this is a serious shot across the bow to any old-style AI startups or businesses which thought “our moat is our data, no one has the millions of datapoints necessary (because everyone knows DL is so sample-inefficient) to compete with us”. The curves here indicate that just training a large as possible model on broad datasets is going to absolutely smash anyone trying to hand-curate finetuning datasets, especially for the small datasets people worry most about...
* eg the text->python example: the text is basically just random Internet text (same as GPT-3), Common Crawl etc, nothing special, not particularly programming-related, and the python is Github Python code; nevertheless, the transfer learning is very impressive: a 10x model size increase in the pretrained ‘text’ model is worth 100x more Github data!”
Philpaper archive sends recommendations of similar articles.
Some notes on Kriorus.
It allows “sign up after death”: that is, a relative may try to sign up for an already deceased person. Many people were cryopreserved this way when their relatives started googling after the death of a person (or a pet).
Last year Kriorus had internal conflict but the attempt to change management seems to fail.
My concern is that fusing experiences may lead to loss of individuality. We could fuse all minds into one simple eternal bliss but its is nit far from death.
One solution is fuse which is not destroying personal identity. Here I assume that “personal identity” is a set of observer-moments which mutually recognise each other a same person.
It is underreporting. There was an analysis of reported deaths, and the picture is grim: It looks like Russia has several times more deaths than was reported and maybe Russia is a world leader. Additional mortality is 338 000 deaths in 2020 (for 140 million population). Almost all my friends had covid.