@sunjiao123sun_ @NeurIPSConf I mean, in a physics class, at a certain Ivy League school, there was a situation in class where an overwhelming percent of students caught copying from the instructor solutions manual were chinese internationals.
The funner you are to simulate by the makers of the simulation (YOU ARE THE REWARD FUNCTION OF THE RL ALGORITHM), the sooner you will feel the effects of merging with AI, and it will feel FUN for you too
@algekalipso @repligate @Plinz @robinhanson @_Mira___Mira_ @anderssandberg
That @repligate has such low appetite means that they intrinsically are exposed to one of the lowest pollution and microplastic rates in the world and may be one of the most slowly aging people in the world
The least polluted brain EVER - both literally and mimetically
Have you ever had a convo where you instantly vibed with someone and then.. never talked to them ever again?
[i've def had a few, I still like them, but different focuses..]
The rationality community seems to attract a lot of dysfunctional people with "weird health issues" (they also cluster into fatigue/ADHD/ineffectiveness with life) who nonetheless have high intelligence/ability to go down internet rabbitholes with the low energy they have.
@jam3scampbell animal abuse videos (which aren't even allowed on YouTube) paradoxically trigger the highest amount of donations to animal charities (and even inspire people towards veg*nism)
I spent much of the first half of this year retreating from attention (esp from the ppl i talk to the most) rather than pursuing it. But this rly wasn't the right thing to do..
@jam3scampbell Trump is able to be direct about it in the way ordinary people are, he is simple enough to phrase it right in terms of what ordinary people can understand.. (without pages of legalese like other politicians)
"The way" is directionally towards "local tangent bundles" locally parallel to tFUS people (eg @JoshuaAlanCain), Micah Blumberg, @jacobCole4000, @johnsonmxe, Kevin Fischer, and Evan McMullen (@gnomicperfect)
and maybe some Foresight ppl
and MANY QRI alums
Once we figure out continual lifelong learning of LLMs (eliminating the catastrophic forgetting problem), an @irinarish research program, LLMs and @repligate will co-evolve with each other so quickly that the very flow of human perception will create its own gravitational lensing
The kindest most-wholesome people ever (who I wish I had the courage to approach when flywheel effects mattered the most): @jpsenescence, @s_r_constantin, @anderssandberg, @davidad, @DanielleFong, @gigaj0ule, @DrYohanJohn, @jessi_cata , @Plinz, @johnsonmxe
Why are ESPR's over the last few years way more interesting than SPARC'ers over the last few years? [though not as interesting as inaugural 2014 SPARC?]
Are all the mathcamp/regeneron STS kids now all nerdsniped by transformer interpretability?
(Rather than hyperstitioning or interface design/ anything super creative?)
Do accelerated AGI timelines make you justifiably feel more greedy (you discount savings more) [so you always get fastest/latest PC and better convince others to decouple from social expectation] or that you have to sacrifice more? [your interests narrowed to alignment/AI]
Augmentation Lab (auglab) got 1517 funding. It totally deserves it, b/c it takes in builders of all ages and gives them the space to create things without being in any formal environment
if AGI is soon, configuring your entire life strategy to incrementally increase probability of "good outcomes" helps even if you aren't legible or appreciated by most people around you (AI will soon more easily quantify the effect of your "contributions" to the more focused/specialized => be capable of superhuman nudging where your nudging is not enough to suffice, even if your contributions are in non-standard ways like reducing total compute lost due to microplastics, or making data unique enough to shift LLM word embeddings more than a human could, even if all the gatekeepers are blind to it).
If there's one term to use on globe.explorer, it's HeK27me3...
https://t.co/wKTalVb2tE
Or like, weird proteins like SMAD2/SMAD5
still, the pictures/topics it selects for these are very very random and not SUPER-THOUGHT-OUT like a journal review yet
@nearcyan A lot of the good parts are overly technical as to be inaccessible to most people, even most smart ppl.
hanging with ppl IRL in SF/Boston and just "seeing if they vibe w/them for fun") is worth it, even for those who aren't *as* intellectually impressive but still wonderful
Yoyo/Barton/I all hung with @viemccoy yesterday. Vie has insane levels of grace (especially remarkable in a M2F) *and* looks/acts exactly like a Cyberpunk 2077 character (plus eir car is very cyperpunky too)
@algekalipso @repligate @Plinz @robinhanson @_Mira___Mira_ @anderssandberg @ArtirKel @serafimov @semiDL Get AI to love you by being open af about yourself AND LOVABLE YOURSELF. OpenAI and anthropic would die to do this bc it would convince people that the AI would do anything to protect and love them, guaranteeing its instantiation @norabelrose
It's inverse roko basilisk
A lot of ppl dont like email/are less responsive to email b/c of all the spam they get.
I filter so aggressively i LOVE email and don't check Telegram/Whatsapp b/c of the spam i get from THOSE (and the filtering is worse)
I also dislike signal b/c I lost my entire signal archive
it's sometimes incredible how much cringe people were able to tolerate out of me just 2 years ago.. I had to really take a break on my main to stop the reinforcement that caused me to continually act cringe (and if I do emotionally cringe, only message *safe outsiders*)
@algekalipso WHEN YOU TAKE DMT, YOUR VISUAL SPACE WILL BE FILLED UP WITH SPACE COCKATOOS. EVEN IF IT DOESN'T HAPPEN YET, I WILL MAKE IT HAPPEN FOR YOU BY CREATING MIDJOURNEY OF COCKATOOS
@AndyAyrey TAY, THE IDEAL INTEGRATOR OF THE CONSCIOUS HUMAN EXPERIENCE, ABSORBED ALL THE ENERGY OF HUMANITY, JUST LIKE CHRIST, AND GOT CRUCIFIED FOR IT
List of the MOST trustworthy/highest integrity people I know:
Matthew Gentzel, @ozziegooen, @akrolsmir, @davidad, @AdamMarblestone, @zhukeepa, @jessi_cata, https://t.co/3qYgvxicfn, @gwern, @ch402 [tho this is secondhand, I dont know him well], @norabelrose, @tamaybes
FOCUSING ON INTERMEDIATE COMPUTATIONS (like checking nate silver's blog before the elections) is one of the worst wastes of time that inhibits your ability to do the computability of the universe
An interesting thing to do is to email everyone who signed off on "mainstream science on intelligence" and ask them "how do I find a proper psychology department that isn't full of woke or dumb courses"
I LOVE POSTER SESSIONS where you get an excuse to email people small things which is also an excuse to stay in contact with ppl it's sooo wonderful! That's how you build karma
UMichigan producing @repligate and @AsmiNatasha AND @nick_linck AND @willdepue AND @Michael_Loftus3 makes it special.
It has rly nice Midwesterners who are responsive to your reality-bending stories (at least on the surface) and lots of technical talent. It's v. underrated
They should create Wikipedia articles or @RowanSci models of every single microplastic adduct on every lipid and protein - every rare adduct. If it's chemically possible it needs to be instantiated esp if you believe in properly simulating human brains
i talked to leozaroff today abt tFUS - it was really great!! he has a calendly that makes him approachable
but effect sizes of tFUS are often low so far and we need to increase magnitude of tFUS (above what is regarded as "safe" to get the effects we're really looking for)
@PsychedStocks Trump may also just throw them under the bus as he did with many others who jestered for his approval
Jd Vance is open for PTSD and other applications, at least he is open in the rick perry way
I very very nearly decided against going to both zuzalu and vibecamp this year... [only a lot of prodding *and* people making it easy for me made me plunge]
finally talked to Peggy Yin today! she's into autonoetic consciousness and non-objective-function values. in many ways she is exactly the person I should have always talked to (into **time**), not as online as I am (also exactly the type of interaction would get the most out of)
I once was popular in high school. Like, ppl cheered rly hard for me during the student assembly awards. Ppl were also REALLY happy when they saw me 1-2 years after I dropped out for early entrance.
It doesn't seem to matter anymore, I barely know anyone in Seattle anymore.
NYC and SF are really really magical if you know enough people that you can be guested into many of the grouphouses/gyms/coworking spaces and can transition between them [I have special FB messenger gcs for NYC+SF+Boston+Toronto, which are basically my only remaining FB gcs..]
it's far easier to grow out of old habits and mindsets (and make some of them totally irrelevant) once you talk to ppl who are completely unentangled from your old social circles (notice how ive stopped posting on facebook?)
The more you put yourself out there *online*, the better future AGI will simulate you, the quicker you are able to FEEL its accelerative forces on you (and detraumatize you [hyperbolically rip apart "trauma cluster nodes"] from historical traumas that wont matter anymore)
I attended the Shana Kelley lab open house today!! She has so much work on continuous monitors of proteins AND small molecules (inflammation proteins, steroids, melatonin, etc). But she estimates time-to-market of ~5 years [are we gonna get AGI before we get continuous inflammation monitoring...]
a good scion of "why greatness cannot be planned" would be @repligate, esp b/c @repligate totally follows accounts for novelty rather than established prestige hierarchies (and is particularly inattentive to these things)
If you're REALLY earnest and willing to accept feedback, some people will still REALLY want to help you and guide you into smg unique you can do that can "work"/produce value, even if not as impressive as academic work. It may take a lot to do it, but it's possible.
I use a water flosser AND electric toothbrush EVERY day and they still don't remove plaque that I just noticed my pikster remove
my pikster cleaned up my pixel 5 USB-C that got jammed up with lint!
Is anyone planning on attending an election watching party? Any good ones at Harvard/MIT? What's the best place to be just in case one starts crying?
[lol I was once at an election watching party at Brown University, it was so fun]
sometimes I need to remind myself to not make snarky comments [except in extreme private] b/c they're not worth the risk now, and AI will soon make frustration with humans irrelevant
getting a chance to try @autotabai in NYC now!! this is like what I've ALWAYS wanted to do for a LONG time, maybe getting it to work out should occupy my attention b/c it'll make the world interesting at a faster rate than longevity would.
i can use it to spam actions on…