1. it’s odd to leap to things like housing markets and consumer debt without considering the demographics of startup employees. i believe your graphs are national averages, so are these employees expected to hold more or less debt relative to average? more or less likely to be homeowners v.s. renters? more or less likely to live in specific regions of the country?

2. the initial shock of covid 3.5 years ago was just massive. i get that it was in many ways transformative and not strictly destructive, but still hypotheticals like “a hundred billion decrease in VC funding” just seem so miniscule in comparison. simultaneously we see how the impacts of a sharp shock got dispersed pretty far across time with covid, and this VC bubble popping isn’t nearly as sharp a shock as we’ve known (call it 18mo, based on those burn rates, vs 2mo over which covid hit the whole country).

3. cascading failures are notoriously difficult to predict. seems to me the real worry is not the title of this post but that the systems which have arrested cascading failures may be eroding. good for bringing up national debt, actually, would be interesting to just embrace this fully and consider food/energy security & geopolitics — but that would make for a pretty different piece.

It seems mostly correct to accept the new calculations in the Improved COTI, which represent a -25% adjustment, and then include the 13% adjustment for taxes, resulting in about a -13% adjustment. This still represents an increase in the cost of thriving.

is COTI actually an inverted measure of the literal “cost of thriving”? i.e. the index goes up when the cost goes down? otherwise, this apparent inverted sign (a -13% change in COTI representing an “increase in the cost of thriving”) is throwing me for a loop.

In broad terms, families with children have seen large reductions in their federal income tax burden, largely due to the introduction and expansion of the child tax credit.

If we want to measure the COTI, as per its original justification, it seems correct to more or less accept the ‘improved COTI’ of -25% instead of -36%, reflecting the errors in health care premiums and college sticker versus effective prices, and various minor fixes. We then must take taxes into account, which should leave us with about a -13% change from 1985 to 2023.

here it happens again: tax costs have decreased since 1985, but somehow that manifests as an increase to the “cost of thriving” index (-25% —> -13%).

To learn gravity, you need additional evidence or context; to learn that the world is 3D, you need to see movement. To understand that movement, you have to understand how light moves, etc. etc.

for the 3d part: either the object of observation needs to move, or the observer needs to move: these are equivalent statements due to symmetry. consider two 2D images taken simultaneously from different points of observation: this provides the same information relevant here as were there to be but 2 images of a moving object from a stationary observer at slightly different moments in time.

in fact then, you don’t need to see movement in order to learn that the world is 3D. making movement a requirement to discover the dimensionality of a space mandates the additional dimension of time: how then could we discover the 4 dimensional space-time without access to some 5th dimensional analog of time? it’s an infinite regress.

similarly, you don’t need to understand the movement of light. certainly, we didn’t for a very long time. you just need to understand the projection from object to image. that’s where the bulk of these axiomatic properties of worldly knowledge reside (assumptions about physics being regular, or whatever else you need so that you can leverage things like induction in your learning).

rationally, automating more tasks in my life should make for an easier life that’s subject to fewer demands. rationally, when this isn’t the case — when individuals each working to automate more things causes them to instead be subjected to more demands (learn new skills, else end up on the street), you shouldn’t expect doubling down on this strategy to be long-term viable.

rationally, if you’re predicting the proportion of people able to stay afloat to be always decreasing up to the singularity — a point at which labor becomes valueless — you shouldn’t expect to still be afloat come that moment.

“rationally”, you’re doomed unless you can slide into a different economic system wherein you do observe the benefits of automation. idly watching your peers get rolled over by that bus is bad for your future as it further separates you from the potential exit ramps. the viable solutions to your problem require collective action. that doesn’t put it entirely out of league with rationality, but if it’s not clear from my tone (i apologize if it reads too strong) i believe you’re thinking of this way too narrowly. i think you’re leaning too far toward a Spock type of rationality for what is increasingly a social problem.

> The tradeoff for connecting with similar people is not connecting with people different from us.

disagree. as you say, micro-communities are aligned very narrowly. which means that if you pair any two random individuals from the same micro-community, they'll be extremely similar along only one particular metric, but randomly different across every metric not relevant to that community. the easiest example of this is nationality: to the degree LW is a micro-community, it connects people of many different nationalities. perhaps the disappointment is that although you're connecting with different people, you aren't connecting over your differences.

> Widespread camaraderie is becoming rarer and rarer. Most of us live in highly polarized societies where the vast majority of people won’t even date across political party lines. More than ever, we should be looking for ways to create camaraderie at scale and soften the dividing impact of micro-communities.

this may be a misleading use of "divide" in the context of "polarization". there's a pretty clear hierarchy here: a societal foundation upon which micro-communities are built. most micro-communities, like ours, aren't antagonistic to other micro-communities nor can they be too antagonistic to the foundation unless they seek literal suicide. in the realm of micro-communities, cross-community disagreements are resolved via distance. but every micro-community is subject to the demands of that massive, volatile foundation. lesswrong.com dies if the nations to which we belong decide they don't want us using their cables to communicate. if not for the power of the macro over the micro, we would not experience the level of division we do today. just look up: is it the micro-communities, or the macro society/state, which wields the power to legislate and enforce every thing which is today the topic of polarization?

to consider your example: i think it's quite right to reject a date from anyone who participates in political action focused on destroying the micro-communities which allow you or i to be as we are. i don't mind different *philosophies*: i'm perfectly happy to have a companion with whom to debate the existence of God -- just so long as they aren't fighting crusades about it. but coupled to every "political belief" is a vote, and behind that, an action in the real world.

during some period of the 18th century, we lived in a politics of construction. legislating things like the Bill of Rights. *creating* choice in religion, choice in speech, choice in which aspects of our life we share with whom (protections from unreasonable “search and seizure”). we created the very protections which allow for micro-identity and micro-communities.

today's is a politics of destruction. it’s about restricting what a person can do to or put in their body; restricting what one can teach to the youth; restricting the very bits and information you and i are allowed to share with each other (not just the tainted label of "free speech", but everything from DRM to intrusive surveillance). the mistake of the day is not that we refuse to date across political boundaries, but that we fail to recognize the very real violence beneath our political abstractions.

> 9/11 broke Americans’ sense of micro-identity. All of a sudden, which team you supported or which political party you agreed with seemed to not matter. Instead, people defaulted to a higher level of identity—being an American.

and this was pivotal to everything you discuss around community. the post-9/11 "camaraderie" led directly to this lasting atmosphere of “if you see something, say something” mutual distrust; to an increased normalization of aggression and dehumanization in our social hierarchies: being fondled by a TSA agent is just a *normal* component of domestic travel; and to the surveillance state which views every micro deviation -- vital to micro-community -- as cause for suspicion.

if this is what political camaraderie creates, then **i don't want that**.

> When the fate of humanity is at risk, we all take on the identity of human.

and of course, EA seeks to promote this type of identity even without the risk. it's also not a terrible place to find those "serendipitous, unscripted, and raw interaction[s]" you treasure. it's especially interesting if you were find camaraderie in any of its spaces... as a collection of growing micro-communities, promoting an ideal of cooperation as it grows toward that dominant macro-community that's today largely void of camaraderie: as EA grows how does the sense of camaraderie change, and if its ideals really did become the universal, would that bring about the type of "at scale" camaraderie you dream of?

> If camaraderie at scale makes ordinary days better, wouldn’t it be awesome to experience it more often? Or at least recreate it on a smaller scale within our own lives and communities. [...]
> What’s missing from the current stack is moments of serendipitous, unscripted, and raw interaction—moments where we find connection in places we least expect it.

i just want to promote in-person conferences and conventions of any kind here. to build on my claim that everyone in this micro-community is meaningfully different in ways we don't display in this medium, IRL gatherings are exactly that opportunity to explore those differences. and especially the nominally entertainment-focused conventions (anything from Comicon to EMFcamp): these specifically create an atmosphere where you can feel both comfortable enough and inspired enough to be creative and spontaneous with complete strangers. the most amazing ones are at their core a chaotic swirling of intensely human passion, some unidentifiable thing that just wants for everyone there to impart a little bit of themselves into whatever's being created and to see eachother in the product. i can't tell you the number of times i've walked past some conference hall at the end of the day, see 4 people painstakingly stacking 1000 chairs, walk inside to help and ten minutes later there's 20 of us and the room's spotless. that alone is satisfying, and it's a short journey from there to far deeper acts of camaraderie.

that camaraderie might not be so immediately focused on "changing the world" or operating "at scale", but it does exist in the moment, it's strong enough to perpetuate itself, and it might provide insights for anyone ambitious enough to recreate it in new environments.

i’d love for anyone to present the argument against this. eq says it’s things like karaoke which make friendships great. the friends i know who are eager to do karaoke are the same ones who will start wild, speculative conversation when we’re idly sitting in the living room together. they’re the interesting people.

the people in my life whom, come the first lull in smalltalk after dinner get uncomfortable and declare “great meal, time to go” instead of opening themselves up for those late-night intimate conversations, are the same people who would turn down an invitation to karaoke.

interesting friends are fun friends. “boring” is the opposite of both “fun” and “interesting”. so if the latter two mean something different to the author than to me, perhaps we agree by saying “build non-boring friendships”?

OpenAI estimated that the energy consumption for training GPT-3 was about 3.14 x 10^17 Joules.

sanity checking this figure: 1 kWh is 1000 x 60 x 60 = 3.6 MJ. then GPT-3 consumed 8.7 x 10^10 kWh. at a very conservative $0.04/kWh, that’s $3.5B just in the power bill — disregarding all the non-power costs (i.e. the overheads of operating a datacenter).

i could believe this number’s within 3 orders of magnitude of truth, which is probably good enough for the point of this article, but i am a little surprised if you just took it 100% at face value.

Should there be an opt-out from A.I. systems? Which ones? When is an opt-out clause a genuine choice, and at what point does it become merely an invitation to recede from society altogether, like saying you can choose not to use the internet or vehicular transport or banking services if you so choose.

the examples given are all networks, with many of the nodes human. if “receding from society” means being less connected with the other humans, then there’s no debate: to opt out of these networks is necessarily to “recede from society”.

but LLMs don’t have this property. they aren’t a medium used to bridge connections between individuals: rather things like chatbots exist explicitly to replace human-human interactions with human-machine interactions, and presently they also serve as knowledge repositories: a single massive node with only one connection, to you, the user. to opt out of this form of human-machine interaction at present is not to recede from society, but rather the opposite.

will this change? surely. but i wouldn’t trust the author’s analogy to be at all useful in understanding how.

Instincts to punish people are how actual humans precommit.

i think you could equally frame this as “people precommit due to an expectation of reciprocity”. like, i don’t generally follow through on my commitments to plans with friends because i fear punishment for breaking them. it’s more that i expect whatever amount i invest into the friendship will be reciprocated (approximately).

you could frame the fallout of a commitment failure as “punishment”, but if the risk of punishment exceeded the benefit of cooperation that would discourage me from pre-committing; from interacting with the thing at all. if i thought my crush would beat me should i break things off with him, then i’d simply never ask him out to begin with and we’d probably both be worse off for that.