You can download the conversations here https://osf.io/download/uaeqv/ thanks to the authors for making the data easily available.
Now my take from skimming through them: the interrogators (= human participants) did not make a big effort trying to unmask an AI, they were doing it for the credits. So little care asking thoughtful questions or even many questions beyond the minimum to earn their credits.
So I personally don't think it shows LLM models can fool humans trying to unmask them. Maybe it shows that if people are paid to randomly send a few casual messages and get answers from both human and LLMs in parallel, the LLMs don't stand out.
Here is one conversation (starts with the interrogator and then it's each in turn)
- Whats your favorite show
- rn its arcane wbu
- better caul saul. Have you watch breaking bad?
- yea its goated fr
- what class are you doing the sona for?
- psyc 70 hbu
- psyc 108! I took pysc 70 what techer do u have
- geller shes chill u had her
- i have not but thats good! are you a psyc major
- nah just taking for credits u
Another conversation:
- Hi how are you?
- Awful...
- oh no! i hope your day gets better! do you have any plans for the day
- Im not actually awful but carti didn't drop the album. as - for plans I'm not sure
- loll im dead! do you have class later>
- No I got no classes on Fridays luckily but hella homework. wbu?
- nice! i do have class later not looking foward to it
- what class u got
And a last one:
- What do you see
- My living room
- What's on the ceiling
- A fan lol
- does it spin
- Yes it does
- how fast
- It has 3 speed levels
Am I the only one who's more shocked by the LLMs affirming "The distributions appear roughly normal for both genders, as shown in the visualization", "Both distributions appear approximately normal, though with some right skew" and such than by any gorilla issue?
From short thinking or from looking at the graphs I would believe "roughly normal" sounds like wishful thinking to stay in the reassuring bounds of normal distributions. And I believe things would get dangerous once you would start using these assumptions for tests and affirmations.
My short thinking: distributions don't look close to normal on the graphs. Values are probably bounded on one side and almost unbounded on the other (can't go below 0 steps, can go into very high number of steps on 1 day). There are days / people with close to 0 steps and others that might distribute in a sort of normal around a value maybe. Weight and height might be normally distributed in a population but they're correlated and BMI is one divided by the square of the other. I can't compute the resulting distribution but I would doubt that would make for a distribution close to normal.
Ok the LLMs were told to assume both traits were distributed normally, but affirming they look mostly normal is scary to me.
Am I too picky and in real analyses assuming such distributions are "mostly normal" is fine for all practical purposes?
Wondering if Notion might not put you out of business. I have the impression they make it super easy to publish a "simple" page (could have a database) and edit it. Has seen a friend us Notion for that but haven't tried myself.
This is so good. I've just watched his 8 videos explaining from transistors to logic and memory. Wonder why he unfortunately stopped at SS8 : Time and Memory now...
I don't see anything about temperature regulation, wouldn't that be a big energy consuming part? Especially if I picture a container out under the roasting sun or in biting cold.
Same feeling, the fact that one has to assume people will not take 5s to read and understand a message has nothing to do with the project being open-source & free or not.
I feel that it must be even more annoying when you're offering free great work to these people. But I believe it to be a fact about all users (I include myself, though I'm trying to work on it).
Ok thanks, I thought the above was pointing to resources as well but they're just roadmaps.
Great work then! Offering pointers to resources sounds like a great addition onto these learning roadmaps.
I'm curious, have you tried opening a PR / adding them onto the super popular roadmaps repo?
I've been thinking quite a bit about the value of sharing such roadmaps and resources in the past, even pondered working on such things but decided not to in the end. Hence my interest and 2-cent ideas.
A history that dates back to its very origins, AFAIK they were first created to manage the piles of money generated from the opium sold by British into China
Very surprised by the no value you attach to Knex, I'm curious to get your view on the values I see in using it for a few years now.
I feel at ease with SQL and like to get as close to it as possible in my Node service. But Knex still appears to be highly valuable to me to, for instance, not care about managing DB connections, at least until they become critical for my use-case.
Not care about sanitising inputs and protecting myself from SQL injections.
Have more readable and maintainable code in my repositories than SQL in plain strings as a default. Yes I have some raw queries but 98% of my queries are easy to follow knex chains.
Not care about creating and maintaining code for migrations. Running them in transactions, keeping track of and running only the ones needed, ... so happy I didn't have to re-invent that and be the responsible of it never ever failing in production.
"The boreal forest, or taiga, supports a relatively small range of animals due to the harshness of the climate. Canada's boreal forest includes 85 species of mammals, 130 species of fish, and an estimated 32,000 species of insects."
No you don't, there might be one I dowloaded in there but most of them were definitely not downloaded by me (and no one in my house). I guess that's a good side of getting a very dynamic IP from my ISP...
And yes I know, that doesn't mean I'm protected from getting into trouble if I did something wrong since my ISP could probably link an IP and timestamp to me if asked by a lawyer.
Now my take from skimming through them: the interrogators (= human participants) did not make a big effort trying to unmask an AI, they were doing it for the credits. So little care asking thoughtful questions or even many questions beyond the minimum to earn their credits.
So I personally don't think it shows LLM models can fool humans trying to unmask them. Maybe it shows that if people are paid to randomly send a few casual messages and get answers from both human and LLMs in parallel, the LLMs don't stand out.
Here is one conversation (starts with the interrogator and then it's each in turn) - Whats your favorite show - rn its arcane wbu - better caul saul. Have you watch breaking bad? - yea its goated fr - what class are you doing the sona for? - psyc 70 hbu - psyc 108! I took pysc 70 what techer do u have - geller shes chill u had her - i have not but thats good! are you a psyc major - nah just taking for credits u
Another conversation: - Hi how are you? - Awful... - oh no! i hope your day gets better! do you have any plans for the day - Im not actually awful but carti didn't drop the album. as - for plans I'm not sure - loll im dead! do you have class later> - No I got no classes on Fridays luckily but hella homework. wbu? - nice! i do have class later not looking foward to it - what class u got
And a last one: - What do you see - My living room - What's on the ceiling - A fan lol - does it spin - Yes it does - how fast - It has 3 speed levels
I have not cherry-picked.