coderatlarge

in a slightly different direction: many sites also take it upon themselves to implement ip blocks. for example, us-based utility companies or healthcare companies seem to block IPs from “strategic competitor” countries (let’s say) but also from many others including supposed allies.

the page also says

“ GNU Artanis was Certificated as Awesome Project at 2013 Lisp in summer projects “

so i guess this is not news?

“When you have four or five layers between the client and the actual worker, each taking a cut, it becomes impossible to track where influence ends and legitimate business begins."

wow i would think that for all the money paid to ERP vendors tracking five layers of contracts would not be a major issue.

but i guess unless people in positions of authority subject themselves to some sort of financial audit then it’s almost impossible to detect corruption related payments.

isn’t detailed information about the user equal to additional billing power? perhaps the only disincentive that exists to having that information would be such overwhelming risk/liability that it would outweigh the profit potential of having it in the first place. it seems to me the relative incentives have reached a oretty stable equilibrium…

maybe we need a law against selective enforcement of laws. together with the comprehensive statistics collection agency that would be required to enforce it.

but why use “~float16” for something that in its own documentation the authors describe as a float16* ?

can it execute the y-combinator?

arguments like this cost anthropic nothing; violating privacy will cost them lawsuits.

it sounds to me like an attempt to shame the user into ceasing and desisting… kind of like how apple’s original stance on scratched iphone screens was that it’s your fault for putting the thing in your pocket therefore you should pay.

extending that line of thought would suggest that anthropic wouldn’t turn off a model if it cost too much to operate which clearly it will do. so minimally it’s an inconsistent stance to hold.

and here is a real-deal medical device literally intended to liquify cells at a distance:

https://pmc.ncbi.nlm.nih.gov/articles/PMC9404673/

i personally have no problem with people including AI gen’d code without attribution so long as they stand by it and own the consequences of what they submit. after all, we all know by now how much cajoling and insisting it takes to get any AI gen’d code to do what it’s actually requested and intended to do.

the only exception being contexts that explicitly prohibit it.

i got my first industry job through interview performance, but in later job transitions my university affiliation ended up playing a major role because certain employers are extremely focused on certain feeder schools, at least for periods of their existence often through founder bias or because it buys a certain kind of harmony based on shared experience and shared world-view.

bluelightai.com blog

interesting analysis of errors on a test set

has this story not been told many times before in scifi icluding gibson’s “neuromancer” and “agency”? agi is when the computers form their own goals and are able to use the api of the world to aggregate their own capital and pursue their objectives wrapped inside webs of corporations and fronts that will enable them to execute within today’s social operating system.

while the model companies all compete on the same benchmarks it seems likely their models will all converge towards similar outcomes unless something really unexpected happens in model space around those limit points…

fwiw, i agree with most of the points in @Fomite’s response below. the people i’ve known fall into a perverse version of his “ego” point.

they felt that when they got tenure they “won” and their “ego” was strong enough to allow them to ignore the disapproval of their peers for not doing the conventionally expected things. they felt that they knew better in their hearts what the discipline truly needed and that the rat-race of establishment approval wasn’t it. so they turned inward. which is not necessarily the healthiest path imo.

true but ignores handing over all your prompt traffic without any real legal protections as sama has pointed out:

[1] https://californiarecorder.com/sam-altman-requires-ai-privil...

i’ve personally known a number of tenured professors who’ve systematically shirked all responsibility after their tenure event. they’ve been willing to live as semi-pariahs within their peer group though.

even when required to teach they simply repeat classes they’ve taught many times before making no effort to optimize for reviews.

i don’t doubt your experience but i wonder how much it has to do with not wanting to endure your colleagues’ and departments’s disapproval vs actual threat to employment.

and fwiw, i’m not saying it has to be this way just that it can be this way due to the structure of the system. similarly there are many corporate situations in which one can scrape by for extended periods of time, but there is rarely a “for life” clause. even so, it hasn’t prevented the university system from helping to catalyze all the amazing discoveries we all benefit from in society every day.

i know of prestigious departments where after literally decades of political stalemate with colleagues (over things as petty as who gets what office) prestigious faculty finally managed to finagle a high-dollar offer from a lower tier institution and de-camped over the politics.

in many countries the salaries are unbelievably low by US standards, but they generally do come with healthcare, benefits and a pension.

maybe those who fight for it have better information.

for example they realize that once they achieve tenure, the amount of work truly required to retain the for-life annuity is risibly low so they can go on to do just about whatever else they want or “consult” for extra dollars as needed.

many academics also seem willing to invite industry people to guest lecture in their classes

understood thanks - but in those few cases how do you even determine who fails?

sure, maybe it was intended as a novel coinage, but i assumed the “vicious” interpretation which is the more common one since the comment explicitly references Sayre's Law.

which students fail your class?

my kid attends a school in which they’ve given up on lectures. each “class” is basically a proctored mini self learning test from a booklet that’s a mix of content and exercises to work through individually. a teacher is around to answer questions and grade the booklets.

many kids fail to make the transition from spoon-feeding to self-learning, but those who do then begin to realize that they can go as fast as they are able and need not follow the herd. they also develop a strong sense of whether they’ve understood each booklet or not. it leads to a competition for learning fast AND well because there are also traditional proctored checkpoint exams from time to time plus kids do the ordinary standardized tests to calibrate.

i feel it’s an excellent system that prioritizes learning over conformity though it is obviously not a candidate for mass adoption because many kids wash out after making no progress for a while.

in my experience, teaching quality does benefit from repetition (it is also harmed by it!).

the way i’ve generally understood the use of the word “vicious” in this context is in judging other academics’ work quality. which is also typically where i think most people from the outside perceive the stakes to be low: as in who cares whether one more journal article that no one will read gets published? but from the inside it can mean the difference between tenure and no tenure (for the young academic vying for it), respect and abject failure, money or no money.

fair point - though it sometimes turns into a fight about how to remove others’ tenure; which is their most prized and valuable possession.