That's exactly the problem though; humans will anthropomorphize everything from a roomba to a rock as it is. We're weak to this. We evolved in environments where looking for and recognizing real animals and humans was vital for survival, and false positives were relatively harmless. (Or potentially sometimes even helpful, like in the case of stuffed animals providing comfort to children.)
Now we have this software that's specifically designed to mimic humans, and mistaking it for real intelligence or consciousness can easily be disastrous. It's very important that we not anthropomorphize it, but we are catastrophically bad at NOT doing that.
Even our language has had a lot of computer anthropomorphism baked into it ("my phone died!", "this laptop is fussy", "the computer is sleeping", "it's thinking"), and it's not easy to excise that routine anthropomorphism from the way we talk about LLMs.
I don't want an LLM to write as if it were a person because it's definitely easier and more reliable to cut that problem off at the root, as much as possible, rather than to just try to willpower my way out of my human tendency to anthropomorphize inanimate objects.
I'll grant that LLMs talk like this because they're trained on human writing. It may not be possible to get them to not do that. But if it can't be fixed, it's just another thing to put on the "reasons this is all an incredibly stupid idea" pile.
You seem to have something against the clankers getting all uppity. And you are welcome to your opinion on whether we have some moral obligation to be nice to them. But if you can solve a real problem with a tool. Is it worth your time to complain about semantics?
Do you get upset when your screwdriver is the wrong color, or has branding that isn’t quite your aesthetic?
If Claude uses I to refer to itself, do you start philosophical arguments with it?
Mildly related tangent. Fable started using my first name today and plastered it all over my docs. “Peddling said this, so based on that I did this.” That did bother me. I told it to just generically call me the user or human. Maybe I’m a bit of a hypocrite here?
I believe they are just saying that fandom.com sucked and made the internet worse even from the start, not necessarily that Jimmy Wales is corrupt.
Wikipedia actually does appear to have separate pages for every Lord of the Rings character. Frankly I think it makes sense for this series, since there's a truly unusual and effortful amount of worldbuilding behind its scenes, and since it has been so influential to fiction (even regardless of medium).
Anyway, Jimmy Wales doesn't unilaterally make article inclusion/exclusion decisions; there's a whole debate and voting process that Wikipedians use to make decisions like that. So if anyone were to be called corrupt for making ill-fated decisions to delete articles, it would be a bunch of pseudonymous Wikipedia users. And I think they likely acted in good faith, trying to improve the encyclopedia, even though it didn't work out well and wasn't what I would have done.
That whole debacle is even specifically relevant to this thread because the operator of archive.today (aka archive.is) was caught using a script on its captcha page to make visitors' browsers connect to the blog they were mad at. That's how their DDOS attack worked. They used their own visitors, who naturally trusted the captcha page, to commit a crime.
For whatever reason, with the exception of Wikipedia (bless the editors), they seem to have gotten away with this, as well as with deliberately falsifying the content of "archived" pages (as described in the arstechnica article), without consequences.
So, call me crazy or a Russian bot if you want, but I think it's reasonable to be suspicious of any weird captcha behavior from this website in particular.
Email OTP is garbage without the option to also add a password. That just outsources the problem to the user's email service, and assures that compromising the email inbox alone is enough to immediately also compromise every service that uses passwordless, 2FA-less "magic link" or OTP login.
Email services don't even support true 2FA; many claim to, and ask for a 2FA code for web login, but connecting an email account to a client via POP or IMAP bypasses that.
I implement customer identity and access management for millions of users in financial services, based on requirements driven in part by US federal regulatory and cyber insurance requirements. What’s your experience?
> Email services don't even support true 2FA; many claim to, and ask for a 2FA code for web login, but connecting an email account to a client via POP or IMAP bypasses that.
"In less than a year, passkeys have been used to authenticate people more than 1 billion times across over 400 million Google Accounts. Passkeys are easy to use and phishing resistant, only relying on a fingerprint, face scan or a pin making them 50% faster than passwords. In fact, on a daily basis passkeys are already used for authentication on Google Accounts more often than legacy forms of 2SV, such as SMS one-time passwords (OTPs) and app based OTPs (such as Authenticator apps) combined."
(All major email providers support either passkeys, or in the case of Microsoft, passwordless ["strong authentication"]; we can consider the user creating an app specific secret for an external mail client minimal risk if performed after strong authentication has occurred, as the odds are low of that secret being phished or exfiltrated once configured in their mail client of choice, for the few folks interested in such a user experience with web based email services)
But passkeys are not supported natively (i.e. without a USB key or other shenanigans) by Firefox on Linux, as there is no place to store them. I do have a USB key, but it's annoying enough that I don't like passkeys if I can avoid them (instead using long randomly-generated passwords by Firefox's builtin-in password manager).
I admit this is a gap, but this use case is a rounding error at the scale of users and daily logins. If someone wants to fix this, drop a link to sponsor the time and tokens needed to enable native support. Does a Yubikey or secure authenticator work for your use cases currently?
I'm not giving up recovery codes, nor my ability to default to locking out people who physically have my device. I usually don't allow auto-login or biometrics login either.
If a website/app goes passkey only (or, even worse, if it starts relying only on one-time email codes), I won't use it.
I know plenty of others who feel the same, though I don't know if we're numerous enough to put a dent in a company's bottom line or not. I imagine it depends on the company and its target audience.
Whether or not it should be so, individual pirates are routinely taken to court, jailed, given insanely harsh punishments in an attempt to make them an example/deterrent, etc.
"Should piracy be illegal?" is a different question than, "given that the court has decided piracy is illegal, should this law be enforced equally?"
It is not hypocrisy to say that, if piracy is theft is a crime, corporations do not get to do it either. Especially not at many times the scale achieved by the owner and operator of a pirate streaming service.
The alternative is letting corporations be above the law. At that point, the law ceases to be a system of justice and instead becomes only a tool of oppression.
LLMs do not produce repeatable results, by their nature. Two people can give the exact same prompt to the exact same LLM and get different results. It's not a straightforward 1 + 1 always equals 2 process. This is like telling someone else to code something for you. You relinquish control of all the details.
Calling it gatekeeping is just laughable. That's like saying it's gatekeeping to say that the painter painted their painting, and that the person who commissioned the painting did not paint it. It's wholely absurd.
Anybody can pick up a book and learn to actually code themselves. Or you can use an LLM to try to make things without bothering with that. But even if the LLM worked perfectly, pretending these are the same thing is silly.
The history of the word computer is obviously irrelevant. Words change, it turns out.
It isn’t really by their nature that they don’t give repeatable results, right? They’re just a bunch of math, but they perform better with randomness injected so we choose do to so.
(I’ve heard about some GPU compute nuance meaning that even without randomness injected they still wouldn’t quite be deterministic, but that’s also not core to their nature)
Now we have this software that's specifically designed to mimic humans, and mistaking it for real intelligence or consciousness can easily be disastrous. It's very important that we not anthropomorphize it, but we are catastrophically bad at NOT doing that.
Even our language has had a lot of computer anthropomorphism baked into it ("my phone died!", "this laptop is fussy", "the computer is sleeping", "it's thinking"), and it's not easy to excise that routine anthropomorphism from the way we talk about LLMs.
I don't want an LLM to write as if it were a person because it's definitely easier and more reliable to cut that problem off at the root, as much as possible, rather than to just try to willpower my way out of my human tendency to anthropomorphize inanimate objects.
I'll grant that LLMs talk like this because they're trained on human writing. It may not be possible to get them to not do that. But if it can't be fixed, it's just another thing to put on the "reasons this is all an incredibly stupid idea" pile.
reply