There used to be industries of artisans making high quality products.
They are still there. You can get hand made clothes, shoes, cars, shotguns, guitar amps, bicycles, food, furniture, art, software, buggy whips, everything. Most people cannot or don't want to pay for them.
Before the industrial revolution, people that weren't rich didn't have high quality products made by artisans. What they had was low quality and inexpensive. The industrial revolution greatly increased both the quality and quantity of products most people could afford.
Artisanal shoes are no longer an industry, they're generally either scattered individuals or assembly line products that are advertised as handmade despite not being particularly high quality.
Regarding quality back in the day, my understanding is that e.g. a farmhand could buy a solid pair of work boots for around a week's wages. These boots would be custom-fit, probably full grain leather, and reliable because everybody knows everybody by name and the cobbler probably has a reputation to maintain.
These days, equivalently custom boots look like they're roughly a day's to a week's wages, but the lower you go in that range the more likely you are to get scammed into buying dropshipped garbage.
Anthropic spent a lot more time editing their pre-training data with higher-quality material. For example, they used actual books for pre-training, because books are generally run through human editors for higher quality. (this is from the copyright lawsuits)
[ ... ]
Meanwhile its obvious that companies like Google would use trash data like email and chat logs or Internet forums for pre-training. Just complete trash data.
Google has been scanning every book they can get their hands on for 20 years or more. They have a bunch of patents on non-destructive book scanning gear. Surely they are using that data to train their models.
Have you ever tried the Gemini models? They are absolutely atrocious. Like, hilariously bad. One time they kept asking me for "Harriet! Where is Harriet?!" and another time it responded in chinese.
They were not trained on good data.
And higher-quality source material results in higher-quality emergent behavior, everything from thinking to tool use.
Seriously people, read their copyright court case. It literally states why Anthropic focused on books and higher quality material.
Google has to learn to not use trash internet content if they want to catch up.
(note that I haven't tried Neon yet so maybe they learned their lesson and fixed it. All my experience is from 3 months on their Ultra plan and previous models from earlier this year)
The logic is: when you have a lot of money, you get a lot of influence and you get the freedom to choose what to do. Be it traffic humans, induce riots, steal property, etc. Oversimplified, but this logic holds most of the time in my experience.
No, the logic is that if you have a copy of a book, then create a copy and destroy it, you still have a copy of a book (and so can't go and sell on the original).
If I recall correctly its because legally they (Anthropic) can avoid a copyright lawsuit if they destroy a rare book, because digitizing it and then having the original no longer exist in any other form is somehow... it prevents any possibility of a lawsuit due to a quirk in the law.
While Anthropic destroyed the books and the fact that the new copies substituted for the old ones was a factor in the fair use analysis, it may not be necessary. Google already prevailed in the case brought against them by the Authors guild for nondestructively scanning for building a search index and only showing snippets to searchers.
In the same review of The God Delusion, Eagleton writes "Nor does [Dawkins] understand that because God is transcendent of us (which is another way of saying that he did not have to bring us about), he is free of any neurotic need for us and wants simply to be allowed to love us."
That doesn't sound like something an atheist would say.
There is a fantastic response to both quotes in the comments section of the LRB article where Eagleton's review was published.
Terry Eagleton charges Richard Dawkins with failing to read theology in formulating his objection to religious belief, and thereby misses the point that when one rejects the premises of a set of views, it is a waste of one’s time to address what is built on those premises (LRB, 19 October). For example, if one concludes on the basis of rational investigation that one’s character and fate are not determined by the arrangement of the planets, stars and galaxies that can be seen from Earth, then one does not waste time comparing classic tropical astrology with sidereal astrology, or either with the Sarjatak system, or any of the three with any other construction placed on the ancient ignorances of our forefathers about the real nature of the heavenly bodies. Religion is exactly the same thing: it is the pre-scientific, rudimentary metaphysics of our forefathers, which (mainly through the natural gullibility of proselytised children, and tragically for the world) survives into the age in which I can send this letter by electronic means.
> That doesn't sound like something an atheist would say.
I can describe what a Flat Earther believes without agreeing with them. Eagleton is simply describing the nature of what classical Christian theology believes is the nature of God: Dawkins is attacking a straw man and Eagleton is simply calling him out on it.
Dawkins's critiques of (straw man) proofs of God's existence ("Everything has a cause.") are howlers. He goes no where near Aristotle's or Leibniz's arguments:
IMHO the real skill is in understanding/explaining your opponents positions as well as (or better than?) he can, and then putting forward arguments against that:
Dawkins built his pop career on taking apart religious beliefs, so this seems like an odd response. He obviously wants to spend time on the beliefs of others. He just chooses to attack the ones that are easy to attack, often focusing on things that are closer to folklore than academic positions of theologians. It's like disproving Santa Claus. Yeah, you can do it, but maybe it's not as profound as it seems.
Theology rests on faith. There is no experimentation, no proof, nothing other than inventing a story and then inventing a theological framework that story fits into, then calling it "academia" because within your invented story, there is a framework you can argue.
> Theology rests on faith. There is no experimentation, no proof […]
Mathematicians have no experimentation. Nor do logicians. And if you're talking about "experimentation", the word you are looking for is "evidence", not "proof".
> It isn't real.
If you're arguing that the only "real" things are ones that can be be empirically measured / verified, then down that road you will find incoherence:
Faith rests on experience. There is experimentation, but it is experimentation with yourself. Imagine an experiment where the success depends on how honest and sincere the experimenter is.
An example of such an experience would be Pascal's "night of fire".
The core of theology is a body of writings about that experience. Not everyone does that; e.g. Pascal kept it private. There are first-hand writings (the author had it), second-hand (the author saw someone who had it), and third-hand (the author read other writings). An outsider cannot tell them apart, of course. E.g. William James ("Varieties of religious experience") got interested in this because he himself had a glimpse.
You seem to have read "academic" as "scientific." I think you're forgetting about classics, philosophy, literature, all of which can get very academic (and even rigorous!) without resting on the scientific method.
I won't call you anything, but I think you're reading too much into a word that, most simply, just means scholars who work at established institutions of higher education.
Do you consider the art department or the literature department at your local university to be "not real" because they don't do scientific experiments?
Yes it is, but it's an awkward term to use, because of invoking "proof" at all.
"Disproving" means refuting, which means testing. It would be nice if the idiomatic phrase was "the burden of testability".
I mean, disproving works fine. If I believe in, say, snatter goblins, you can challenge me by saying "snatter goblins are supposed to emerge from brickwork at night, but careful observations have never revealed them, therefore there aren't any", and I can respond "they are only visible to those who believe", and you can say "that's an untestable theory then, so it's pretty much false, get out of here," and everything is as it should be.
But if you inadvertantly encourage me to look for proof, I can do things like producing a bent twig and saying "this twig was bent in exactly the way that would have happened if a snatter goblin stepped on it!", or "a witness swears to having heard the tentative cries of many snatter goblins, one night when the moon was full!" or "this unusual musty aroma, perceived by many people at sites where brickwork is present, is the aroma of snatter goblins!" and lots of other useless observations that threaten to lead to lengthy investigations that would anyway be inconclusive, because there's no such thing as proof outside of rigorously logical formal domains, which is to say, proof is a wrong-headed concept in most domains. That's why I don't like the phrase.
"Extraordinary claims require extraordinary evidence" is another way to put it, but suffers from the same problem: getting your opponent to come back with a big bag of supposedly great evidence isn't what you wanted. What you really want to see is a big bag of simple hypotheses that fail when tested. Yet it is an important thing to say.
> moved to iocaine which is slightly different in what it does, and that was fine until so many scrapers hit my site that it saturated my upload with generated content feeding the them
You're feeding bots poison without rate limiting them all to some small fraction of your pipe?
iocaine has shipped firewall support to temporarily drop traffic from IPs after you've fed them an appropriate amount poison. More people should enable it.
You should not traffic shape the egress lest you run out of workers handling legitimate traffic. You should also make sure connections are appropriately terminated on your side for the same reason.
reply