I am baffled that, in the year 2026, I am telling technically competent engineers to not solicit nor take legal advice from a random word generator
-
I am baffled that, in the year 2026, I am telling technically competent engineers to not solicit nor take legal advice from a random word generator
@Possiblydrew People I had known to be very clever have fallen to these. I have a pet hypothesis that there's a factor unrelated to general intelligence that makes some people more susceptible to these than others, like gambling addictions.
-
I am baffled that, in the year 2026, I am telling technically competent engineers to not solicit nor take legal advice from a random word generator
@Possiblydrew stupidity and competency are technically different
-
I am baffled that, in the year 2026, I am telling technically competent engineers to not solicit nor take legal advice from a random word generator
Engineers never had a good grasp on what technology actually did: they're mostly formula followers
-
@Possiblydrew Have you actually tried the current Coding assist AIs?
Because if you had, you would be surprised they are not "random word generators" but actually very useful.
You can try https://antigravity.google for free on your own code on your own machine (Linux is fine) and ask it to do your own choice of tasks.
If you want to come back and tell me you found rand() in an overcoat, well, OK. But I think you will be very surprised.
@hopeless @Possiblydrew So which of your favorite LLM tools are deterministic?
-
@hopeless @Possiblydrew So which of your favorite LLM tools are deterministic?
The LLMs are actually deterministic themselves; you can give them the same PRNG seed and get the same result.
It's wrong to say that they are "random word generators" like there is a dictionary and a random index into the dictionary is chosen... they are choosing what they want to talk about based on the weights they arrived at.
You too will not recite what you want to say verbatim each time, but through your internal state decide how to express yourself.
-
The LLMs are actually deterministic themselves; you can give them the same PRNG seed and get the same result.
It's wrong to say that they are "random word generators" like there is a dictionary and a random index into the dictionary is chosen... they are choosing what they want to talk about based on the weights they arrived at.
You too will not recite what you want to say verbatim each time, but through your internal state decide how to express yourself.
@hopeless wow there’s so much to unpack here.
Random + weights + other stuff is still
random . Random + seed sure is deterministic but how often do you the user set the seed? Why is the answer to an objective query subject to a probabilistic state?
-
I am baffled that, in the year 2026, I am telling technically competent engineers to not solicit nor take legal advice from a random word generator
@Possiblydrew if they're doing that, then they most definitely are not technically competent in any way whatsoever.
-
I am baffled that, in the year 2026, I am telling technically competent engineers to not solicit nor take legal advice from a random word generator
@Possiblydrew Apropos of nothing, taxes are due tomorrow.
-
I am baffled that, in the year 2026, I am telling technically competent engineers to not solicit nor take legal advice from a random word generator
Ok, but also there's this: https://www.harvey.ai/ but I don't think they license to individuals
-
The LLMs are actually deterministic themselves; you can give them the same PRNG seed and get the same result.
It's wrong to say that they are "random word generators" like there is a dictionary and a random index into the dictionary is chosen... they are choosing what they want to talk about based on the weights they arrived at.
You too will not recite what you want to say verbatim each time, but through your internal state decide how to express yourself.
@hopeless @hastar @Possiblydrew
they are choosing what they want to talk about
Wow, what a coincidence. I was just talking to /dev/urandom and it chose to talk about "bizz".
-
@hopeless @hastar @Possiblydrew
they are choosing what they want to talk about
Wow, what a coincidence. I was just talking to /dev/urandom and it chose to talk about "bizz".
@barubary @hastar @Possiblydrew
You're so clever!
-
The LLMs are actually deterministic themselves; you can give them the same PRNG seed and get the same result.
It's wrong to say that they are "random word generators" like there is a dictionary and a random index into the dictionary is chosen... they are choosing what they want to talk about based on the weights they arrived at.
You too will not recite what you want to say verbatim each time, but through your internal state decide how to express yourself.
@hopeless @hastar @Possiblydrew They aren't sentient, they are not 'choosing' anything. They're just picking at random from weighted options. They don't know anything. If the training set was good for what specifically you wanted to do you might be lucky. Otherwise you could be totally fucked and not know it until a lot later when something breaks/lets someone into your system.
-
I am baffled that, in the year 2026, I am telling technically competent engineers to not solicit nor take legal advice from a random word generator
@Possiblydrew indoctrinating engineers and engineering types with goofy shit is like shooting fish in a barrel. on average they think they're smarter than they are, at least in part because so much education rewards rationalizing over anything else
