No no no it's fine! You get the word shuffler to deshuffle the—eloquently—shuffled paragraphs back into nice and tidy bullet points. And I have an idea! You could get an LLM to add metadata to the email to preserve the original bullet points, so the recipient LLM has extra interpolation room to choose to ignore the original list, but keep the—much more correct and eloquent, and with much better emphasis—hallucinated ones.
And also closing with:
Nvidia insists that it “wins on merit, as reflected in our benchmark results and value to customers.” And Nvidia does have the best stuff — but that’s not what the DOJ, Warren, or France are concerned about, is it?
To tie the bow nicely.
Quinn enters the dark and cold forest, crossing the threshold, an omnipresent sense of foreboding permeates the air, before being killed by a grue.
“Once we get AGI, we’ll turn the crank one more time—or two or three more times—and AI systems will become superhuman—vastly superhuman. They will become qualitatively smarter than you or I, much smarter, perhaps similar to how you or I are qualitatively smarter than an elementary schooler. “
Also this doesn't give enough credit to gradeschoolers. I certainly don't think I am much smarter (if at all) than when I was a kid. Don't these people remember being children? Do they think intelligence is limited to speaking fancy, and/or having the tools to solve specific problems? I'm not sure if it's me being the weird one, to me growing up is not about becoming smarter, it's more about gaining perspective, that is vital, but actual intelligence/personhood is a pre-requisite for perspective.
Hi, I'm going to be that OTHER guy:
Thank god not all dictionaries are prescriptivists and simply reflect the natural usage: Cambridge dictionary: Beg the question
On a side rant "begging the question" is a terrible name for this bias, and the very wikipedia page you've been so kind to offer provides the much more transparent "assuming the conclusion".
If you absolutely wanted to translate from the original latin/greek (petitio principii/τὸ ἐν ἀρχῇ αἰτεῖσθαι): "beginning with an ask", where ask = assumption of the premise. [Which happens to also be more transparent]
Just because we've inherited terrible translations does not mean we should seek to perpetuate them though sheer cultural inertia, and much less chastise others when using the much more natural meaning of the words "beg the question". [I have to wonder if begging here is somehow a corruption of "begin" but I can't find sources to back this up, and don't want to waste too much time looking]
I feel mildly better, thanks.
Meanwhile some of the comments are downright terrifying, also the whole "research" output is overly-detailed yet lacking any substance, and deeply deeply in fantasy land, but all the comments a debating in favour of or against what is perceived as "real work", and in terms of presentation "vibes".
I mean my parents always said that fascist/cultish movements have issues distinguishing signified and signifier, but good grief. (Yes too much Lacan in the household)
And yet they can spit out copyrighted material verbatim, or near-verbatim, how strange and peculiar.
Not every rationalist I've met has been nice or smart ^^.
I think it's hard to grow up in our society, without harboring a kernel of fascism in our hearts, it's easy to fall into the constantly sold "everything would work better if we just put the right people in charge". With varying definitions of who the "right people" are:
Do they deserve better? Absolutely, but you can't remove their agency, they ultimately chose this. The world is messy and broken, it's fine not to make too much peace with that, but you have to ponder your ends and your means more thoughtfully than a lot of EAs/Rationalists do. Falling prey to magical thinking is a choice, and/or a bias you can overcome (Which I find extremely ironic given the bias correction advertising in Rationalists spheres)
It makes you wonder about the specifics:
Maybe they could have solved context issues and possible latency issues by seating the workers in the cars, and for extra quick intervention speed put them in the driver's seat. Revolutionary. (Shamelessly stealing adam something's joke format about trains)
Student: I wish I could find a copy of one of those AIs that will actually expose to you the human-psychology models they learned to predict exactly what humans would say next, instead of telling us only things about ourselves that they predict we're comfortable hearing. I wish I could ask it what the hell people were thinking back then.
I think this part conveys the root insanity of Yud, failing to understand that language is a co-operative game between humans, that have to trust in common shared lived experiences, to believe the message was conveyed successfully.
But noooooooo, magic AI can extract all the possible meanings, and internal states of all possible speakers in all possible situations from textual descriptions alone: because: ✨bayes✨
The fact that such a (LLM based) system would almost certainly not be optimal for any conceivable loss function / training set pair seems to completely elude him.
Transcribed:
Max Tegmark (@tegmark):
No, LLM's aren't mere stochastic parrots: Llama-2 contains a detailed model of the world, quite literally! We even discover a "longitude neuron"Wes Gurnee (@wesg52):
Do language models have an internal world model? A sense of time? At multiple spatiotemporal scales?
In a new paper with @tegmark we provide evidence that they do by finding a literal map of the world inside the activations of Llama-2! [image with colorful dots on a map]
With this dastardly deliberate simplification of what it means to have a world model, we've been struck a mortal blow in our skepticism towards LLMs; we have no choice but to convert surely!
(*) Asterisk:
Not an actual literal map, what they really mean to say is that they've trained "linear probes" (it's own mini-model) on the activation layers, for a bunch of inputs, and minimizing loss for latitude and longitude (and/or time, blah blah).
And yes from the activations you can get a fuzzy distribution of lat,long on a map, and yes they've been able to isolated individual "neurons" that seem to correlate in activation with latitude and longitude. (frankly not being able to find one would have been surprising to me, this doesn't mean LLM's aren't just big statistical machines, in this case being trained with data containing literal lat,long tuples for cities in particular)
It's a neat visualization and result but it is sort of comically missing the point
Bonus sneers from @emilymbender:
~~Brawndo~~ Blockchain has got what ~~plants~~ LLMs crave, it's got ~~electrolytes~~ ledgers.
@ESYudkowsky: Remember when you were a kid and thought you might have psychic powers, so you dealt yourself face-down playing cards and tried to guess whether they were red or black, and recorded your accuracy rate over several batches of tries?
|
And then remember how you had absolutely no idea to do stats at that age, so you stayed confused for a while longer?
Apologies for the usage of the japanese; but it is a very apt description: https://en.wikipedia.org/wiki/Chūnibyō,
Not surprised, still very disappointed, I feel sick.