Scimitar
9mo ago
Why are gays over-represented in the arts and creative fields? Even in tech, I go to an artsy coding meetup and it's hosted in an LGBT space. I go to a discord of people building interesting things and they're 50% furries.
I would presume there's a biological/psychological explanation, if the effect is even real, but it's hard to find good answers. My first guess would be the same loosening of priors that allows for creativity also loosens the heterosexuality prior; but then why gay and not bi?
Scimitar
3mo ago
https://old.reddit.com/r/madlads/s/yV09R82gov
I came across this reddit post of a guy who cuts his books in half. I found it funny that this triggered people in the comments. Many people refuse to fold pages or write in their books. I don't see what the big deal is. To me a book is just a tool. I understand taking care to preserve semi-rare books, but these were books you can just walk into any bookstore and buy.
I used to know someone who would even try to avoid creasing the spine. But I've also heard about a guy who would rip pages of shlocky fiction as he read them to keep his place. Where are you on the spectrum?
Scimitar
4mo ago
New Year's resolutions? Do you make them? Do you keep them?
This year, I am thinking of the following priorities:
- Work hard(er). Get into the habit of hard work.
- Continue to meet new people, and maintain existing relationships
I came across The Top Five Regrets of the Dying, and I agree with all of them but "I wish I hadn't worked so hard." I actually wish I was working harder and really would like to maintain a consistent daily work habit. I managed this a couple of month this year but it dropped off at the start of October. I should probably read Atomic Habits even though I feel like I will know most of it.
I have more ideas floating in my head but I won't set them as goals just yet. I should tidy up and publish a crappy project I have, just to get it out there. I should release more work in general, or build in public. I want to perform in front of an audience. I should figure out how to leverage the internet to make friends online. I should get into climbing. I want to go on a solo trip somewhere. These are more specific and I might pick some of these later to focus on, but I know for sure those two priorities above define the general direction I want to go.
Scimitar
6mo ago
Does anyone have access to ChatGPT with images? Is there some opt-in setting I have to find and enable, or are they still doing a staged roll-out?
Scimitar
5mo ago
This is a lot of words and all quite complicated. In my experience it is a "simple" habit modification.
You have an established habit of negative self talk that has become second nature.
I'm running late for the dentist. Why am I so useless at everything?
The first step is to just become more aware of this reflex.
I'm running late for the dentist. Why am I so useless at everything? Oh, there's the negative self-talk reflex.
Once you are aware of it, you can practice reframing or rephrasing, and talk to yourself like you would a friend.
I'm running late for the dentist. Why am I so useless at everything? Oh that's negative self talk. Actually it's no big deal. People are sometimes late for things and I was busy with work.
And then you can stop yourself as you have the negative response
I'm running late for the dentist. Why am I so -- actually it's not a big deal. I'm good.
And then you just jump to the healthy response
I'm running late for the dentist. It's no big deal. Can't wait to get dinner afterwards.
Each step is building a habit, and like any habit requires a bit of effort and mindfulness to start with, but gets easier with time. It shouldn't take years, probably closer to a week or two for each stage. I probably isn't Nick Cammarata MDMA-fuelled self love, but is a good way of getting you out of these automatic negative thought patterns.
Scimitar
1yr ago
A few months ago OpenAI dropped their API price, from $0.06/1000 tokens for their best model, to $0.02/1000 tokens. This week, the company released their ChatGPT API which uses their "gpt-3.5-turbo" model, apparently the best one yet, for the price of $0.002/1000 tokens. Yes, an order of magnitude cheaper. I don't quite understand the pricing, and OpenAI themselves say: "Because gpt-3.5-turbo performs at a similar capability to text-davinci-003 but at 10% the price per token, we recommend gpt-3.5-turbo for most use cases." In less than a year, the OpenAI models have not only improved, but become 30 times cheaper. What does this mean?
A human thinks at roughly 800 words per minute. We could debate this all day, but it won’t really effect the math. A word is about 1.33 tokens. This means that a human, working diligently 40 hour weeks for a year, fully engaged, could produce about: 52 * 40 * 60 * 800 * 1.33 = 132 million tokens per year of thought. This would cost $264 out of ChatGPT.
https://old.reddit.com/r/singularity/comments/11fn0td/the_implications_of_chatgpts_api_cost/
...or about $0.13 per hour. Yes technically it overlooks the fact that OpenAI charge for both input and output tokens, but this is still cheap and the line is trending downwards.
Full time minimum wage is ~$20k/year. GPT-3.5-turbo is 100x cheaper and vastly outperforms the average minimum wage worker at certain tasks. I dunno, this just feels crazy. And no, I wont apologize for AI posting. It is simply the most interesting thing happening right now.
Scimitar
1yr ago
Do you think Greta actually writes her own tweets? I can't imagine an autistic Swedish teenager tweeting like that.
Scimitar
1yr ago
Am I supposed to feel good after exercise?
People say things like "Every time I run I feel like anxiety doesn’t exist for awhile and just feel good." or "Does anyone know exactly why it feels so fucking good to lift heavy shit above your head?? I have never felt anything like it in a gym. It is THE MOST addicting feeling I’ve possibly ever had." both quotes from random reddit comments.
I've tried both running and weight lifting in the past, and never got a huge positive psychological effect. Mostly I just feel slightly content, or just plain tired. Do some people just not have a strong positive response?
Scimitar
3mo ago
I can make a "real person" custom GPT, but what kind of questions would you ask it exactly?
How do you keep track of what you've read online? How do you manage your personal research? Do you have a system? Do you have collections, and keep notes? Is this an actual problem you have?
Personally, I often read something and then struggle to find it again months later, or open something in a new tab to read later only to lose the tab. Or I might write down my thoughts but then lose the context of what I was looking at when I had them. It just feels like things here could be much better.
Scimitar
10mo ago
Does anyone have experience of fixing poor forward-neck posture? I'm happy to do 10 mins of daily stretching, as long as the stretching will actually work. Or is it fixed by doing something else? What exactly is the mechanism?
I do facepulls at the gym, which may have helped my rounded shoulders, but my cervical spine still gives me away as a nerd.
Scimitar
1yr ago
What's your definition of AGI? The label feels more like a vibe than anything else.
For me, the multi-modal capabilities of GPT-4 and others [1][2] start to push it over the edge.
One possible threshold are Bongard problems[3]. A year ago I thought, while GPT-3 was very impressive, we were still a long way from AI solving a puzzle like this (what rule defines the two groups?) [4]. But now it seems GPT-4 has a good shot, and if not 4, then perhaps 4.5. As far as I know, no one has actually tried this yet.
So what other vibe checks are there? Wikipedia offers some ideas[5]
-
Turing test - GPT3 passes this IMO
-
Coffee test - can it enter an unknown house and make a coffee? Palm-E[1] is getting there
-
Student test - can it pass classes and get a degree? Yes, if the GPT-4 paper is to believed
Yes, current models can't really 'learn' after training, they can't see outside their context window, they have no memory... but these issues don't seem to be holding them back.
Maybe you want your AGIs to have 'agency' or 'conciousness'? I'd prefer mine didn't, for safety reasons, but would guess you could simulate it by continuously/recursively prompting GPT to generate a train of thought.
[1] https://ai.googleblog.com/2023/03/palm-e-embodied-multimodal-language.html
[2] https://arxiv.org/pdf/2302.14045.pdf
[3] https://metarationality.com/bongard-meta-rationality
[4] https://metarationality.com/images/metarationality/bp199.gif
[5] https://en.wikipedia.org/wiki/Artificial_general_intelligence#Tests_for_testing_human-level_AGI
Scimitar
1yr ago
What are the current usage limits? Does the history work? What is the maximum input length for a single comment (presume 8k tokens)?
Scimitar
1yr ago
Scott Aaronson describes a feeling (that I too experienced):
Here’s the thing: I spent my formative years—basically, from the age of 12 until my mid-20s—feeling not “entitled,” not “privileged,” but terrified. I was terrified that one of my female classmates would somehow find out that I sexually desired her, and that the instant she did, I would be scorned, laughed at, called a creep and a weirdo, maybe even expelled from school or sent to prison.
I was not as bad as Aaronson, but I held a completely unquestioned conviction that no girl must ever know how much I desired her, and that no one must know I had any sexual thoughts at all. I have no idea where it came from, but it seemed as evidently true to me as the fact that the sky is blue.
So, reading this story, I wonder if this innate impulse is actually adaptive for spergs. If you don't understand the social landscape of romance and dating, then indeed your best bet is to opt out and hide. If you try to play without understanding the rules, you end up ostracized or worse.
Here's a short conversation. But it still claims it's not conscious; maybe it's all the RLHF.
Scimitar
4mo ago
Don't fall for the passion meme
Scimitar
5mo ago
Back to medicine, as long as I have pharmaceutically enhanced diligence, I've yet to run into anything I simply can't understand, and I think I do a decent enough job at it.
Medicine is not massively g-loaded, is it? It's mostly memorisation, hence the popular med school anki subreddit. I can't think of anything day-to-day that actually requires reasoning about. I think it's why doctors suck at thinking from first principles and shut down when you ask them about something outside their "training set".
Scimitar
8mo ago
In the UK, a 19 y/o was found guilty of a hate crime for quoting a rap lyric containing the n-word on her instagram page. A while ago, a man was jailed for 20 weeks for posting 10 "grossly offensive" George Floyd memes in a private Whatsapp group. Another man was found guilty of sending an offensive tweet, celebrating the death of Sir Tom Moore, and the tweet was only live for 20 minutes. Recently a man was found guilty of wearing an offensive football shirt (a reference to the Hillsborough disaster over 30 years ago). There are many more examples in kind, these are just from the top of my head.
I think these are all absurd. Anything that offends normie public sensibilities is illegal. I would much prefer the American free speech norms.
Scimitar
1yr ago
I found an comparison between davinci-003 and chatgpt. Seems like chatgpt is better at some things and worse at others; not a blanket downgrade: https://scale.com/blog/chatgpt-vs-davinci.
Also there is no more data flywheel:
Data submitted through the API is no longer used for service improvements (including model training) unless the organization opts in
How have they cut costs? I've seen answers I don't understand. It doesn't seem like they are necessarily running at a loss:
Quantizing to mixed int8/int4 - 70% hardware reduction and 3x speed increase compared to float16 with essentially no loss in quality.
A*.3/3 = 10% of the cost.
Switch from quadratic to memory efficient attention. 10x-20x increase in batch size.
So we are talking it taking about 1% of the resources and a 10x price reduction - they should be 90% more profitable compared to when they introduced GPT-3.
I did some quick calculation. We know the number of floating point operations per token for inference is approximately twice the number of parameters(175B). Assuming they use 16 bit floating point, and have 50% of peak efficiency, A100 could do 300 trillion flop/s(peak 624[0]). 1 hour of A100 gives openAI $0.002/ktok * (300,000/175/2/1000)ktok/sec * 3600=$6.1 back. Public price per A100 is $2.25 for one year reservation.
https://news.ycombinator.com/item?id=34986915
Essentially, I had previously thought that better language models will need to be orders of magnitude larger and orders of magnitude more expensive. Now it seems clearer that we're moving towards a future of intelligence too cheap to meter. And "endless stupidity generator"? I just plain disagree, and feel compelled to build a thin client product on top of the API to prove it. There are many jobs that are mostly low-skill word manipulation, and now that you might get that 100x or 1000x cheaper, it opens up yet unimagined opportunities IMO.
Scimitar
1yr ago
What meaningful difference is there between a fetus and a newborn? Neither can survive independently and neither has a sense of self. Killing one is like killing another.
There is some fuzzy boundary when a zygote becomes a "person", but I would argue it happens far enough after birth for it not to make a difference here.
Scimitar
4mo ago
It's a nice checkpoint to consider the last year and strategize about the next.
Scimitar
5mo ago
Indifference isn't always a "healthy" response, but it can be. Every situation is different and there are many good and bad ways to think about each one.
Now instead of thinking about whatever bad event set me off, I'm thinking about how stupid and irrational and unhealthy and undisciplined I am for engaging in negative self-talk.
I was going to mention this but didn't want to make things too complicated. You are allowed to have the negative self talk, but then you have to practice the rephrasing even if you don't feel it. Would you talk to your friend like that? How would you think about things if it was your friend and not you? And actually think it "out loud".
On negativity - I argue negativity is unhealthy. I know what you mean, it's no good thinking everything is great when it's not, but you can still maintain good vibes while acknowledging and fixing mistakes.
Scimitar
6mo ago
If I marry an aspie woman, what are the odds that we get profoundly autistic kids? Does anyone have an intuition on this, or any data?
Scimitar
2mo ago
Yeah, I learned to code.
Why do to want to switch careers? What's the issue with your current one? It is important to understand yourself. I found working a simple service sector job far more tiring than programming, because it takes me a lot of energy to talk to normies all day. I didn't understand this about myself for a while, but it would have helped when choosing a career.
I don't trust any "career councilor" type person. The best way to learn about a career is to talk to someone who does it. In retrospect, I should have gone to meetups much earlier to get to know people and the industry. I would suggest doing that for careers you are interested in.
This is already the case, but now everyone complains about food waste. Can't win!
More options
Context Copy link