Rendered at 23:56:45 GMT+0000 (Coordinated Universal Time) with Cloudflare Workers.
asen 3 hours ago [-]
fun thing is that of course it will forget it in a long session and spit some too long text. but just typing `dirac` makes it recall the instruction and shorten it
markus_zhang 2 days ago [-]
I actually take this advice in real life. I don’t talk unless really needed. There are exceptions like to my kid, but that’s the general principle to non family/friend members.
frangonf 8 hours ago [-]
To me adding 'tldr' at beginning or end of the prompt whenever I need is enough. Sometimes I want more explanatory responses.
codegladiator 2 days ago [-]
- Only Use Attempto Controlled English
- Do not make any sentence in negative phrasing
asen 2 days ago [-]
Thanks. And maybe I am wrong, but I feel that Paul Dirac should be better represented in the weights than Attempto Controlled English, which sounds niche to me - or at least - I hear it for the first time.
Easy for anyone to try out themselves, and my post is only half a joke - it actually works. For me, at least :)
FrankWilhoit 2 days ago [-]
You want people to talk to you the way you listen. So does everyone else, and most of them do not listen the way you do.
asen 2 days ago [-]
this is an instruction for LLM, not for people
FrankWilhoit 2 days ago [-]
LLMs are what you get when you put people through a stack of filters.
pixelpoet 2 days ago [-]
Dirac would never make that less vs fewer mistake.
asen 2 days ago [-]
Agreed, thanks! This kind of makes it obvious that I am not a native speaker, but kind of also proves its was not written by an LLM :)
Edit: fixed now: less words -> fewer words
dcminter 1 days ago [-]
FYI it's a somewhat regional or (in the UK) class stratified grammar distinction.
While nobody would say "fewer sand", "less" is in practice often used by native English speakers for both count and non-count nouns so it's not really wrong so much as it is revealing.
mejutoco 5 hours ago [-]
> is in practice often used by native English speakers for both count and non-count nouns
Alternative explanation: native speakers are wrong sometimes.
It really boils down to:
- Less + uncountable noun (money, energy, code, sand)
If instead we prefer to attribute it to class and say using the correct or wrong way is a class signifier and that both are correct, and that language depends on usage and there are no grammar rules really but only consensus, we can also do that. But then "Why waste time say lot word when few word do trick?".
I prefer the first explanation.
P.S. Same with "if I were" turning into "if I was". I guess usage turned that one, and that is fine too. I make that mistake often.
There is no English Grammar Institute to tell us what is right and wrong, so English grammar is necessarily descriptive and not prescriptive.
It may annoy you, and you may be convinced that you are right and others are "wrong" but if a substantial body of people use English in a way that you dislike then it's just tough shit I'm afraid.
> But then "Why waste time say lot word when few word do trick?".
Because nobody actually talks like that, that's why.
You don't speak Elizabethan English, or Regency English, or late Saxon, or Old English, but you presumably don't think those are "wrong" - why do you think Estuary English is wrong; could it be gasp because you're in a different linguistic group? /me clutches pearls...
aborsy 2 days ago [-]
Would Dirac get a job these days?
asen 2 days ago [-]
The reason i believe is good is because with very few tokens you encode two properties, for which Paul Dirac is most famous for.
* a great scientist (which hopefully encourages researching and verifying answers)
* a person who doesn't like to use many words at all
The later hopefully compensates the apparent "output-prose-greediness" most models seem to have. No idea if that "token greedines" (or slop) is a result from the companies business models or inherent for the models and their training :/
Easy for anyone to try out themselves, and my post is only half a joke - it actually works. For me, at least :)
Edit: fixed now: less words -> fewer words
While nobody would say "fewer sand", "less" is in practice often used by native English speakers for both count and non-count nouns so it's not really wrong so much as it is revealing.
Alternative explanation: native speakers are wrong sometimes.
It really boils down to:
- Less + uncountable noun (money, energy, code, sand)
- fewer + countable noun (trees, cars, grains, volts, coins)
If instead we prefer to attribute it to class and say using the correct or wrong way is a class signifier and that both are correct, and that language depends on usage and there are no grammar rules really but only consensus, we can also do that. But then "Why waste time say lot word when few word do trick?".
I prefer the first explanation.
P.S. Same with "if I were" turning into "if I was". I guess usage turned that one, and that is fine too. I make that mistake often.
In Spanish these guys (https://www.rae.es/duda-linguistica/el-nombre-de-la-y-es-i-g...) also officially changed the name of "y" (i griega) to "ye". The argument: it is easier for new speakers. I will say "i griega" forever I am afraid.
It may annoy you, and you may be convinced that you are right and others are "wrong" but if a substantial body of people use English in a way that you dislike then it's just tough shit I'm afraid.
> But then "Why waste time say lot word when few word do trick?".
Because nobody actually talks like that, that's why.
You don't speak Elizabethan English, or Regency English, or late Saxon, or Old English, but you presumably don't think those are "wrong" - why do you think Estuary English is wrong; could it be gasp because you're in a different linguistic group? /me clutches pearls...
* a great scientist (which hopefully encourages researching and verifying answers)
* a person who doesn't like to use many words at all
The later hopefully compensates the apparent "output-prose-greediness" most models seem to have. No idea if that "token greedines" (or slop) is a result from the companies business models or inherent for the models and their training :/