All of Borasko's Comments + Replies

Changing my life in 2021, halfway through

I have not, but I'll look into it. I hope it goes well for you!

Big picture of phasic dopamine

Very interesting! I don't know much about the brain but I think this post did a good job of explaining the concept and showing it's importance. I wonder how the brain does this with neuroplasticity.  I've read this article from MIT about researchers rewiring eye inputs to the audio processing parts of the brain. Would the hypothalamus have that hyper-prior on what eye data "looks like" and create loops and systems that could de-code that data and reintegrate it with the undamaged processing systems? Could an AI system just as easily create or re-use e... (read more)

Is driving worth the risk?

I don't want to die, but I also want to live. The future is inherently uncertain, so if I were to take a decreased quality of life (no more driving) to better my chances of surviving to ASI, I had better have strong intuition that ASI would come. 

I'm short on AI timelines (2035 median, right tailed distribution), but I also drive and take more risky activities on covid because I like being in person with my friends and family. The fact that I don't know what a post singularity world would look like helps me feel comfortable taking these risks, it seem... (read more)

3adamzerner1moYeah, I definitely hear ya. I have these feelings too. But at the same time, I think it's in violation of Shut Up and Multiply [https://www.lesswrong.com/tag/shut-up-and-multiply]. I hear ya here too. It's once of the main places that affects the conclusion, I think. My reasoning for expecting a post-singularity year to have positive utility is because it seems like a place where it'd make sense to adopt the opinion of the experts I quoted in the article. (And that it's less depressing.)
AMA: Paul Christiano, alignment researcher

If you believe AGI will be created. What would be the median year you think it will be created at? 

Ex. -2046, 2074, etc. 

2065

That's an estimate for TAI (i.e. world doubling every 4 years), not sure what "AGI" means exactly.

Broad distribution in both directions, reasonably good chance by 2040 (maybe 25%)?

Don't hold me to that. I think it's literally not the same as the last time someone asked in this AMA, inconsistencies preserved to give a sense for stability.

Borasko's Shortform

I think I am approaching the end game of video gaming.

 I can scroll through Steam and GoG for hours while seeing nothing that slightly interests me. Every once and a while a great game comes and revolutionizes my experience with video games, but like riding a roller coaster its then hard to go back to the Farris wheel after and feel a rush, all games after are discounted on their novelty. The games that did that for me were the Pathologic "series" and Disco Elysium.  Yesterday was a bit different, I went on a site called Legendsworld.com and they... (read more)

4Viliam2moAre you familiar with eXoDOS [http://exodos.the-eye.us/]? It's approximately 100GB torrent of 7000 games for MSDOS, configured to run out of the box -- the torrent includes a few clones of DOSBox and similar software, each game already comes with a configuration that works, also manuals and maps, you just download the huge thing, start the UI and play. I used to play computer games when I was younger; now I have less time to play, and also it is less fun when I cannot fully focus on the game because there are all kinds of worries and interruptions that come with having a job and kids. So I can no longer be like "I am going to play this game until it is completed, even if it takes the entire weekend". But yes, there is also the part of "I have already seen most of it". This sounds likely. Internet is insanely addictive. When you play an old game now, it also misses the social aspect: no one else you know is playing it at the same time. So you can't discuss the game, can't compete, etc. Winning at something other people also care about was part of the fun. But I think that at least 90% of the reason is internet. The vast amount of alternatives it provides (even if most of them are meh) makes it difficult to focus on one specific thing. Also, too many alternatives make it unlikely that you and your friends would choose the same thing.
Convict Conditioning Book Review

It's brushed over in the book but he believes weights put unnatural stress on the body causing it to deteriorate quicker. Little evidence other than his anecdotal experiences is given for this. It is mostly the assert and move on.

Covid 4/9: Another Vaccine Passport Objection

I work at my county's health department helping with the vaccination process and I can confirm that you are scheduled for two weeks out exactly when your first shot is done.  If you miss that date or have to reschedule you are put on a waiting list for a different date in time,  I don't have access to the list, but I also assume it's scheduled for you again with the hopes things will just work out.

I take inbound calls sometimes and many people are worried if they cancel they won't ever get, or since it's been longer than 2 weeks "something" would... (read more)

Covid 4/1: Vaccine Passports

As far as QR code privacy goes, anybody that has their location checked on in their phone is getting tracked much more than that code could ever do. Although I'm sure they wouldn't mind having the 'vaccinated/non-vacinated' categories. If being vaccinated wasn't already inferable by reading the persons emails / listening to their phone calls.

I really like the idea of privacy. But I just assume after Snowden's leaks everybody is being maximally tracked 24/7. I'm not saying its right, but I have no power to stop it. So when I hear of people's privacy concern issues about being tracked the best I can do is be sympathetically supportive, but I believe it's a lost cause and thus usually pre-factor it in to government proposals.

1tkpwaeub2moPrivacy concerns could be addressed by periodically deleting data more than ten days old.
Politics is way too meta

"But the front-page articles really shouldn't be about the controversy, the buzz, the second-order perceptions and spin and perceptions-of-spin."

 

I totally agree, but this seems like a business optimization problem. Considering all the different news sources that compete against each other for their consumers very limited time, they have to make it click-baity and scandalous if they want people to pick them over a competitor. I think sometimes when we get really close to an election that the election basically does become a tv show or a sports game. I... (read more)

There's a certain perception of what "respectable journalism" looks like, and this perception is what causes the New York Times and CNN to not immediately rush down the slope to tabloid journalism in pursuit of short-term clicks.

I think this "respectable journalism" image affects newspapers' behavior because the public has this concept in mind, and many people will consume news less if it seems too far from respectable journalism. Separately, this image also affects newspapers' behavior because the journalists care about "respectable journalism" to some de... (read more)

Borasko's Shortform

That's a good idea, I'll try that.

Fun with +12 OOMs of Compute

Awesome! Thanks for your answer!

Fun with +12 OOMs of Compute

 This was really interesting to read. I'm still pretty new to the AI space so I don't know how this compares to our current FLOP usage. Assuming our current course of computing power doesn't change, how long is the timeline to get to 10^34 FLOP of computing power? 

6Daniel Kokotajlo3moWell, it's 12 OOMs more than our current FLOP usage. ;) Ajeya's report is excellent and contains the best and most thorough answers to your questions I think. If I recall correctly, she projects that the price of compute will drop in half every 2.5 years on average, and that algorithmic/efficiency improvements will have a similar effect. And then there's a one-time boost of up to +5 OOMs that we get from just spending a lot more. So she projects we get to (the equivalent of) +12 OOMs around 2050. Personally, I think the price of compute will drop a bit faster in the next ten years, and the algorithmic improvements will be a bit better too in the near term. But I'm very uncertain about that and mostly just defer to her judgment.
Borasko's Shortform

At the start of this year I stopped playing video games except when in social situations. My hope for doing this was I would be able to study more without the distraction, and sometimes playing videogames encourages behavior that lead me to be more reclusive than I think is healthy for me.

This worked fine for the first two months, however the last two days have been really rough on my mental state. I found myself breaking down and playing video games last weekend. I was bummed. I planned to go a full year without doing that. But considering I am still aliv... (read more)

3John_Maxwell3molsuser [https://www.lesswrong.com/users/lsusr] had an interesting idea of creating a new Youtube account and explicitly training the recommendation system to recommend particular videos (in his case, music): https://www.lesswrong.com/posts/wQnJ4ZBEbwE9BwCa3/personal-experiment-one-year-without-junk-media [https://www.lesswrong.com/posts/wQnJ4ZBEbwE9BwCa3/personal-experiment-one-year-without-junk-media] I guess you could also do it for Youtube channels which are informative & entertaining, e.g. CGP Grey and Kings & Generals. I believe studies have found that laughter tends to be rejuvenating, so optimizing for videos you think are funny is another idea.
2MikkW3mo>I'm proposing I play [video games] for a fixed amount of time per day as a trade. I'll suggest doing this as a fixed amount of time on a weekly, not daily basis- some days the quota you set may be too high (because you've gotten really into some productive thing that you want to do more than play video games), other days it may be too low (because you just need time to chill out), and having a weekly quota of video games lets you adjust according to your mood on any given day.
Borasko's Shortform

Those are good to know! Thanks!

Borasko's Shortform

I seem to be going back into a major depressive state, emotionally it doesn't feel much different from my normal functioning depression, but I notice a largely decreased ability to devote mental energy and get started on tasks. Words seem to not make sense, like everybody is writing in Wernicke's Aphasia, it's all there and fluent but the words have no meaning to me.

When I try to code or do mathematics my ability do calm down and try to solve problems get severely reduced, I short circuit to irritability and anger that I can't quickly solve what I thi... (read more)

5jimrandomh4moThis seems self-aware and accurate, and means you have a decent chance of being able to intervene and avoid going into a depressive state. The two things that sound particularly high leverage (and I believe tend to feature prominently in standard advice, because they tend to be key elements of self-reinforcing spirals) are sleep and diet. https://www.lesswrong.com/posts/a6PMaSrfG9KYWL9tL/how-to-improve-your-sleep. [https://www.lesswrong.com/posts/a6PMaSrfG9KYWL9tL/how-to-improve-your-sleep.] Especially the bit about checking for sleep apnea, if you haven't already. Eating more sweets/carbs is a common depression failure mode, and people tend to respond by deciding that the carbs are bad, and trying to stop themselves from eating them. This is a mistake because what's actually going on is undereating of everything else; sweets make up the gap because they're very convenient, but the right intervention is to add more high-quality food until there's no appetite left for sweets.
Making Vaccine

Thanks for posting this, I didn't know about radvac before and now I am excited. I will read the white paper and probably make some myself. If your results are good which I really hope they are, I will try to help my family members get some as well. I don't mind paying $1000+ for family safety, and with the delayed vaccine rollout I would feel better even getting the vaccine to them one month earlier.  So thanks again! I hope all goes well.

The GameStop Situation: Simplified

I haven't seen almost any traders going off a "real value" analysis for Game Stop. Almost everybody believes Game Stop has a broken business model with no fundamentals, but are all buying it and taking losses just to screw over hedge funds. This is coordinated short-sited financial shitposting out of spite. There is bound to be many losers, but man is it interesting to watch.

 

Edit: I would also love to see an analysis at one point of the game theory involved getting so many individual traders to coordinate.

It doesn't feel like game theory to me as much as psychology at this point. During the endgame, I agree game theory will play a huge role in individual investors' decisions to hold or sell. But now, it feels like a clear recognition that they are stronger together as long as they all hold, and what's most interesting to me are the simple slogans and short repeatable talking points they use to do so (as well as the excellent analysis that has gotten them this far).

"WE LIKE THE STOCK"

"IF HE'S STILL IN, I'M STILL IN"

"DIAMOND HANDS"

Why do they work so well? Be... (read more)

Group house norms really do seem toxic to many people.

I lived in a Fraternity for most of my undergraduate schooling. The same problems you had we also had. noise, cleanliness, politics, amount of clothing worn in common areas, taking up space in common areas, showers, money. Except, fraternity level. It seemed every semester there would be at least a few altercations between roommates, it's just natural. 

However we being an 'organization' really helped us function as a group of a bunch different people living in one house under one banner. We specifically had internal structures for dealing with grievan... (read more)

Babble Thread

{I don't know how long these can be but this might be a long post, mostly a vent on frustrations with online none university learning}

1.

When I search for a new skill to learn there are hundreds and hundreds of tutorials and websites and courses that are instantly presented. Some of them seem more high quality than others, which is easy to tell both from websites hosting them (Khan Academy, etc), and then the ratings those individual courses have inside the websites. But there is also free vs paid courses. As somebody who has tried many open online learning... (read more)

DALL-E by OpenAI

I wonder if something like this could be pared with AI Dungeon? If they do release a image generator model for public or private use I think it would be fun to see an image accompany the last line(s) of the text output that has been generated for the story thus far. 

Then more complex AI generated games wouldn't be too far away either.

3Azai A5moTaking a sentence output by AI Dungeon and feeding it into DALL-E is totally possible (if and when the DALL-E source code becomes available). I'm not sure how much money it would cost. DALL-E has about 7% of the parameters that the biggest model of GPT-3 has, though I doubt AI Dungeon uses the biggest model. Generating an entire image with DALL-E means predicting 1024 tokens/codewords, whereas predicting text is at most 1 token per letter. All in all, it seems financially plausible. I think it would be fun to see the results too. What seems tricky to me is that a story can be much more complex than the 256-token text input that DALL-E accepts. Suppose the last sentence of the story is "He picks up the rock." This input fits into DALL-E easily, but is very ambiguous. "He" might be illustrated by DALL-E as any arbitrary male figure, even though in the story, "He" refers to a very specific character. ("The rock" is similarly ambiguous. And there are more ambiguities, such as the location and the time of day that the scene takes place in.) If you scan back a couple of lines, you may find that "He" refers to a character called Fredrick. His name is not immediately useful for determining what he looks like, but knowing his name, we can now look through the entire story to find descriptions of him. Perhaps Fredrick was introduced in the first chapter as a farmer, but became a royal knight in Chapter 3 after an act of heroism. Whether Fredrick is currently wearing his armor might depend on the last few hundred words, and what his armor looks like was probably described in Chapter 3. Whether his hair is red could depend on the first few hundred words. But maybe in the middle of the story, a curse turned his hair into a bunch of worms. All this is to say that to correctly interpret a sentence in a story, you potentially have to read the entire story. Trying to summarize the story could help, but can only go so far. Every paragraph of the story could contain facts about the
Blog plant

I've been interesting in getting into plant life lately. Do you have any learning resources you would recommend?

2KatjaGrace6moI actually know very little about my plants at present, so cannot help you.
Quick Thoughts on Immoral Mazes

I think the rise of the popularity of the internet has actually strengthened moral mazes in companies. 

Especially now since everybody on the internet can access their own very intense small scale subgroups which usually perfectly conforms to whatever ideology / life style that person wants. Not only does the internet allow people to find their niche interests, it allows them to jump between group cultures relatively easily.  In each of these groups there is norms and unwritten culture rules about what is and isn't acceptable, I think that's why t... (read more)

4ChristianKl6moGroups on the internet are not intense small scale subgroups. Joining an organization like Scientology in magnitude more intense then what you get with an internet subgroup. Conflicts inside a company are usually much more important for the people inside the company and their career then whatever online discussions arrise.