• 1 Post
  • 27 Comments
Joined 3 years ago
cake
Cake day: June 1st, 2023

help-circle


  • Interestingly, normies who argue against nationalization of private property tend to say “commies want to nationalize my toothbrush”, etc. not understanding the difference between private property (data centers) and personal property (consumer-grade RAM, for instance).

    So, funny enough, Altman is advocating for the cartoon villain version of communism where the centralized authority (Sam Altman) nationalizes your personal property (RAM).

    Meanwhile, folks actually in favor of nationalization to reap the benefits of planned systems of production would not try to steal your RAM, but would take Altman’s datacenters for the public good.


  • Definitely not. It’d not been nearly as competitive, Musk and other republican donors would not have needed to donated $100M or whatever they’re up to now, and the Democrats would just do their normal thing (and lose)

    But, it’s Paxton- the 2nd most corrupt politician in America - so they had to divert funds from other close races and grassroots donations for Talarico have broken records. And the Republicans are still (probably) going to lose.





  • They won’t lowball you below range, though. So it is a “competitive” salary in that sense. It almost always behooves you to ask for more than you reasonably think they would give you.

    Also, usually recruiters are judged almost exclusively on “time to fill”, meaning they aren’t judged by how cheaply they get people in the door, just how fast they do it.


  • I have some insight into this.

    It actually means that:

    1. companies identify peer groups of companies that they compete with for hiring,

    2. they get aggregated data on everyone’s pay from data brokers (they submit anonymized compensation info, brokers aggregate it and distribute the aggregate data to their customers)

    3. they pick a market percentile of the compensation range of their peers that they target for various jobs

    4. they build pay ranges around that target percentile

    Effectively, if most companies target pay “at market” i.e. market median, it normalizes industry pay to the median. This does depress wages, but also provides upward pressure at the low end. Sort of like a price fixing scheme.

    Some companies target higher than that, and those effectively provide upward labor market pressure.

    So, “competitive” means “sort of near to or higher than the statistical median for this job among companies we consider peers”. Unless they’re just lying or misinformed - but they’d have a hard time retaining employees.





  • Funny enough, though their jobs would be the easiest to replace, they are “important” in that they can be held legally responsible for business decisions (Responsible Corporate Officer doctrine). That’s partly why they get tens of millions of dollars - so they can take the fall when shit hits the fan (either they fuck the shareholders or majorly break the law).

    This is known as the “Scapegoating Premium” - you are basically paid for the risk of taking on a very public, high liability/visibility position. The corporation pays their legal fees and has insurance to help with this, but it’s compensation for the possibility of legal and reputational damage.



  • Say you had a version of chat GPT that hadn’t been trained on a corpus that included enough information about the number of Rs the word “strawberry” contains, didn’t have workarounds for character counting, and has context sharing.

    You ask it “How many Rs does Strawberry contain?”

    It burns a bunch of tokens, returns “Strawberry has 7 Rs. No wait, that’s not right, strawberry has 5 Rs. No wait…” etc.

    You say “Strawberry has 3 Rs. How many Rs does Strawberry have?”

    It says “Strawberry has 3 Rs.”

    You ask “How many Rs does this exact string contain: ‘Strawrberrrry’?”

    A new session using shared context responds “Strawberry has 3 Rs.”

    You say “No, it has a different number of Rs. I didn’t ask about Strawberry this time”

    It responds “Sorry about that! You didn’t ask about Strawberry, you asked about Strawrberrrry. Strawrberrrry has 7 Rs. No wait, that’s not right. Strawrberrrry has 5 Rs…”

    LLMs, at their core, use models that have computed lexical & semantic similarity to predict words (really they perform contextualized vector transformations, but let’snot get into that). When they “learn”, they just do this more effectively using more text or using the same text more efficiently. Next word prediction is not the same thing as learning concepts, like the concept of numbers, non-numerical things being ascribed numerical values (ironically, since LLMs function by turning words into numerical vectors), or counting.

    When it responds “Strawberry has 3 Rs” it isn’t because it knows what that means or how to gain that knowledge about other words. It is merely parroting back what you’ve told it because “Strawberry has 3 Rs” in its shared context has very close lexical similarity to your query “How many Rs does Strawberry have?”. It also parses Strawberry as [“straw”, “berry”] encoded into numerical values representing their relationship in the corpus’ vector space (e.g. {[ .420, -.67, .67], [.420, .69, -.69]}) - so unless it has instructions to further break those tokens into characters, it does not have the capability to count. You can actually do some weird math using these vectors and their relationships (their distance apart in the 3D vector space, angles between vectors, etc actually relate to the semantic content of the tokens), but you lose granularity such as the number of letters in a token when you look at words this way.

    Learning is more than just computing lexical similarity. Modern LLMs mimic reasoning by expanding the query with intermediary tokens, basically creating a temporary scratch pad of related words on the same vector, but don’t actually reason (this is known as chain of thought). Learning is partly computing lexical similarity, but it is also about extrapolating concepts from facts and inferences, applying those concepts to novel simuli, etc. LLMs really don’t do reasoning, which is part of learning.




  • We shouldn’t engage in this, to be sure. But it does serve as good propaganda for homophobes(I think), if we’re to find a silver lining.

    If suddenly their political identity is associated with “self-hating gay person”, and they desperately do not want to be perceived that way, they have to distance themselves from that political identity somewhat or defend themselves which will invariably look desperate and keep the story going.

    Similarly, if they know they aren’t using Grindr at these conventions, but they believe there is massive Grindr traffic, then they have to assume tons of people around them are using Grindr - which breeds distrust and ties back into point 1 a bit.

    We still shouldn’t do this - if and when this is widely known to be false, it’ll make the liberals/leftists who traffic in this misinfo, and by extension their views, less trustworthy. I just think that, if you look at it myopically, it’s a good op.


  • Oh there’s definitely a taste difference - mainly because of diet. Farm-raised salmon eat feed made of various grain-starches, oils, and fish meal supplemented by vitamins and minerals while salmon in the wild eat a much different diet. But nutritionally there isn’t a substantial difference.

    As an aside, the salmon fisheries in North America are some of the best-managed and most sustainable in the world. So if you can afford it, then wild caught salmon is a great choice.