Because physically speaking, chaotic and unpredictable are two different things - and why it works so well on this case: it’s becoming a stochastic problem, not a deterministic one.
It’s an awesome area for machine learning: you didn’t need to understand the result and how it got created, it just needs to be “close enough”.
Oh I completely agree with that, just the jump to “a flawed model leaked” is too far. There’s already enough crap to mock, no need to make up additional stuff.
It’s bullshit. What leaked was their commandline tool source code (named “claude code”) - very juicy in itself but has nothing to do with their models.
If you are truly serious: this is not how hallucinations work. It’s really best to think of it as “fancy auto complete”. The hallucinations happen when the next token is too disconnected from what we as humans would call as “belonging together”. But it’s all math after all.
Limit the k value, tube down temperature and cut off context size and the issue of hallucinations is a non-topic for “transcribe and summarize”.
You get into what I’d call “stupid” territory like you’re describing.
Your second point I fully agree with and is the reason why I’d ask the doc directly. To give the personal anecdote: the transcript itself helps me to focus on exactly the topics you’ve described: who’s confused? Where was agreement? Where did people just not speak up?
A specific hallucination example I see every other day for example are tasks: that thing “thinks” that “we should” or “you must” are always tasks and outcomes which is utter bullshit - but I know that and using the transcript part helps me focus on the important part, the humans.
Counter top the popular opinion here for me it would be a clear yes in the situation you’re describing.
The relationship with my direction doc is more believable to me than principles of vague bad feelings for me.
Now taking and transcripts specifically are one of the use cases I also draw value out myself. I’d ask the doc though how they’re using it.
Still I’d rather have my transcript public than to go on that search for a doctor match again.
Again: depends on the legal system that statement is not that easy in that generic form.
Ik specifically thinking about the situation in which a human no longer can communicate what they want:
In Germany there is the legal concept of transferring this kind of decision power in case you yourself are no longer able to do that. The “pulling the plug” situation: the individual can no longer state their wishes directly concerning the situation but left a binding document who has the decision power in their stead.
Now you could argue that this also is “the human has to want it” as they wanted the other person to have this life and death power.
As someone else said: helping humans find a dignified death is legal in some countries.
Your second point is more complicated though: I don’t know the laws in a lot of countries but where I’m from animals are strictly treated as property - emotional connection isn’t taken into strong consideration at all when it comes to assessing their value when it comes to legal fights but they are treated like a distinct thing different from both humans and objects in a lot of other cases (e.g. dedicated laws like “unnecessary” animal cruelty is forbidden ).
About the reason you can discuss as much as you want, the two arguments I’ve stumbled across are:
there must not be a distinction in terms of value because that value must be purely subjective and cannot be assessed.
There is no objective way to classify animals based on emotional connection and therefore the law can’t create categories.
Culturally we treat animals like different to humans all the time - even your dog is not treated “family” to the extreme a child would (think of child protection laws and what that would mean if they’d apply to a dog or a hamster). And now expand this to find a definition which covers both a cow someone has as a beloved pet or a meat animal.
Note that I’m trying to not say wether this is “right” or “wrong”: morale categories and laws have some overlap but they are quite lose as soon as you get specific.
My primary source was an interview with a judge who went into an hour long discussion about how complex the relation between animals and the law is and how “emotional connection” and the need for the law to be objective and repeatable are an inherent contradiction.
In short:
It’s a very tough question because there isn’t the one correct answer. Law, morality and personal subjectivity collide and make a mess out of us.
I’m too disconnected to be knowledgeable enough to recommend or discourage alternatives, just want to note that one has to educate themselves on what is right - your key words will help with that for sure! For example: Just from looking up DoxyPEP: That’s an antibiotic based product and would be useless against HIV. But might be good at reducing the risk of syphilis and other nasty bacterial bullshit.
The other thing where I absolutely agree with you is the issue of the stigma: It must not be a taboo to enjoy oneself, explore own sexuality and be self confident into oneself as a full, human being with whatever sexuality floats ones boat.
That’s why I’m so happy about OP asking even though it’s a shitty situation: I’m sure jt will help some passive reader to realize something about their own sexuality and that’s an awesome thing!
In short: Thank you :)
Edit: sorry for the reply spam I got an error message about posting but it seemed to got through anyway.
To support what others have said: even precum can get you pregnant.
It’s not that “pull out is a tad unreliable” - it is horrible and the failure quota is ridiculously high. And that’s only pregnancy, not even considering the whole disease fun you’re signing up for. With a bit of bad luck that’s a lifetime of annoying medication, no a few weeks.
This means from my point of fire in order of urgency:
A) First of all: stranger, feel loved 🤗 I find it awesome that you have he courage to ask this question!
B) get plan B. It’ll suck if you’re reacting but that’s only a week or two.
C) Review your sexual habits: if you’re into spontaneous drunk sex figure out a reliable protection that you don’t have to think too much about. If it’s with strangers there’s no way around a condom in addition if you don’t want to take something for your hepathitis or HIV for the rest of your life. If it’s people you know and trust then something “just” against pregnancy is enough. If it’s a mix of both I suggest getting a baseline protection that you don’t have to think about in the heat of the moment and condoms as preventative plan B for the cases where you’re not too sure.
It sounds like a cliche but it’s true: The more the dude resists protection the more you’ll need it. After all it will be you who’ll be alone with the consequences. You’re worth more than a few fun moments for some dick!
That’s my problem: any single word humanizes the tool in my opinion. Iperhaps something like “stochastic debris” comes close but there’s no chance to counter the common force of pop culture, Corp speak a and humanities talent to see humanoid behavior everywhere but each other. :(
Accepting concepts like “right” and “wrong” gives those tools way too much credit, basically following the AI narrative of the corporations behind them. They can only be used about the output but not the tool itself.
To be precise:
LLMs can’t be right or wrong because the way they work has no link to any reality - it’s stochastics, not evaluation. I also don’t like the term halluzination for the same reason. It’s simply a too high temperature setting jumping into a closeby but unrelated vector set.
Why this is an important distinction: Arguing that an LLM is wrong is arguing on the ground of ChatGPT and the likes: It’s then a “oh but wen make them better!” And their marketing departments overjoy.
To take your calculator analogy: like these tools do have floating point errors which are inherent to those tools wrong outputs are a dore part of LLMs.
We can minimize that but then they automatically use part of their function. This limitation is way stronger on LLMs than limiting a calculator to 16 digits after the comma though…
Yeah this is beyond ridiculous to blame anything or anyone else.
I mean accidently letting lose an autonomous non-tested non-guarailed tool in my dev environment… Well tough luck, shit, something for a good post mortem to learn from.
Having an infrastructure that allowed a single actor to cause this damage? This shouldn’t even be possible for a malicious human from within the system this easily.
That’s an utterly ignorant statement.
To expect others, often volunteer, to take such a personal risk because the legislation in one part of the world is utterly fucked. How about expecting the people who actually live in the country and state and have a chance to influence those laws to step up their game instead of trying to tell third parties to take individual and personal consequence.
They outline the issues from their perspective.
What else should they do? Break their own licence model (which prohibits (geographic) discrimination) or break the law? It’s either one of those two or comply.
The answer is a clear yes.
In short: Choose your tool that will suit you throughout your degree and really dig into it and learn it now while doing your paper.
Long version:
This is absolutely common and I’m not aware of a text editor which supports footnotes but doesn’t support automatically numbering and referencing.
In latex there’s actually a \footnote that takes care of that. In libre office, if I recall correctly, it’s Insert -> Footnote and I’m sure there are templates with the proper formatting and font sizing already in place.
Now it sounds like you’re quite early in your higher degree career - depending on your goals and future challenges you might want to either go the easiest route or really dig into writing-based formatting: It’s just faster if you’re typing all the time to not switch to a mouse to inert footnotes - but only if your really used to it.
Haha I feel you!
Perhaps try to be a bit less extreme in your language in front of your parents - or offer them a bet: if a CO2 monitor is yellow or red at the evening they pay for it and you can air whenever it’s getting yellow. If it stays green you pay for it and shut up about the windows!
Your parents are right in terms of energy waste, im sorry to say.
I’m quite similar though and air about five times a day but with a timer so that the air gets circulated completely without cooling out the house itself.
The breathing struggle is most likely a mix of dust sensitivity and psychology. For me an air cleaner made a huge difference in my comfort level.
I’d you want to go the scientific approach get a CO2 monitor (the most important element for brain performance when talking about home air) and a PM2.5 sensor for dust levels.
In case you wanna give it a shot: I gave writing samples of myself from chat and emails to a self hosted LLM, telling it to extract the writing style deviations, key elements, common phrases, symbols, patterns, etc. Then gave that as a “answer it this style” system prompt expansion - works like … Quite okay. Still need to go over it or course but it doesn’t sound like marketing bullshit but conveys what I want.
Completely agree with your general assessment though! They’re getting better but the marketing machinery is crazy in their claims.