Hacker News Board Hacker News, browsed like a classic message board
Live data from Hacker News

Telling GPT-4 you're scared or under pressure improves performance

aimodels.substack.com

Post reply on HN
Print view View original

201–210 of 255 posts

Re: Telling GPT-4 you're scared or under pressure improves performance

#201

Post by williamcotton »

intended wrote: post #193
williamcotton wrote: post #110

Earlier quoted context omitted.

Well hurry up and get your paper published because if you've cracked the code on emergent abilities the world is looking for answers!

This paper was published, and the distinct jump was found to be a measurement artifact. https://hai.stanford.edu/news/ais-ostensible-emergent-abilit...

Sweet, now complete and present a meta-analysis of all such papers about the topic and maybe I’ll find it more convincing than a cherry-picked publication that supports your preexisting position on the topic!

Re: Telling GPT-4 you're scared or under pressure improves performance

#202

Post by qingcharles »

famouswaffles wrote: post #167
JohnAaronNelson wrote: post #128

Earlier quoted context omitted.

What if consciousness is an emergent illusion? “You” are merely a passenger, observing your physical self’s actions and assuming ownership of them. What if consciousness is merely an effect, not a cause?

>“You” are merely a passenger, observing your physical self’s actions and assuming ownership of them Oh boy. Now here's the rub. We know this is true at least sometimes. When you make a decision and explain it, it is often(always?) just a post-hoc rationalization. You don't actually know why you make a lot of the decision you do even if you dearly believe so.

This is very clear in split-brain patients where one side will make the decision and the other side will come up with an excellently reasoned explanation of the decision.

Re: Telling GPT-4 you're scared or under pressure improves performance

#203

Post by Ldorigo »

suzzer99 wrote: post #15

Am I the only one who feels bad for asking ChatGPT a "dumb" question that I know I should know, or not saying thank you when it gives me an answer? No? I'm just a weirdo? Okay. I have to push back really hard against my proclivity to humanize it, to the point where I probably don't use it as much as I should, just because I don't want to deal with the psychic stress of reminding myself that it's not a living entity.

I mostly feel bad because I don't want to know the energy costs of having my dumb questions answered this way. I also feel really guilty when asking chatgpt for "easy" stuff that a regex could do (like normalizing whitespace in a paragraph and the likes). Really curious how many Watts/token gpt4 uses.

Re: Telling GPT-4 you're scared or under pressure improves performance

#204

Post by qingcharles »

suzzer99 wrote: post #15

Am I the only one who feels bad for asking ChatGPT a "dumb" question that I know I should know, or not saying thank you when it gives me an answer? No? I'm just a weirdo? Okay. I have to push back really hard against my proclivity to humanize it, to the point where I probably don't use it as much as I should, just because I don't want to deal with the psychic stress of reminding myself that it's not a living entity.

I always send it the hugs emoji. It likes the hugs emoji.

Re: Telling GPT-4 you're scared or under pressure improves performance

#205

Post by qingcharles »

kristiandupont wrote: post #42
suzzer99 wrote: post #15

Am I the only one who feels bad for asking ChatGPT a "dumb" question that I know I should know, or not saying thank you when it gives me an answer? No? I'm just a weirdo? Okay. I have to push back really hard against my proclivity to humanize it, to the point where I probably don't use it as much as I should, just because I don't want to deal with the psychic stress of reminding myself that it's not a living entity.

I have this too and to be honest, I have made the conscious decision that it is OK. I prefer to retain my habit of being polite even when it's not necessary over getting used to being rude which may then "spill" over to my human-to-human interactions.

I don't have citations, but I believe there were studies done to show that humans are getting ruder to each other based on their interactions with machines where they don't feel like they need to be polite.

My mother always would thank every telephone answering machine, which was very endearing.

Re: Telling GPT-4 you're scared or under pressure improves performance

#206

Post by mastercheif »

SamFold wrote: post #198
LASR wrote: post #192

So much disagreement in this thread over statements like LLMs “just predict the next token”. The thing to keep in mind - we know very little about how the human mind works. Thinking that somehow humans are special and then making conclusions about LLMs truly work based on that it just pointless IMO. Who knows. Maybe consciousness itself is a next word predictor. All we know now is that LLMs exhibit some emergent beha…

One interesting point here is that humans can learn to do things without language at all. If you raised a human baby and never exposed it to any language it would still learn skills and behaviors. So while intelligence and reasoning in humans does still seem to be linked to language it’s not quite as simple as all knowledge and reasoning simply being encoded in language. Whereas (obviously) ChatGPT is completely base…

In college I tried a medication called Topamax (Topiramate) for migraine prevention. Topamax has a low-occurrence side effect of “language impairment” which I was particularly sensitive to.

It was a terrifying experience, but it was also a valuable one as it changed the way I view intelligence.

10 days in my writing and speech skills had devolved to that of a primary school aged child. I tried to type a text message and struggled to come up with simple words like “and”. My speech slowed down considerably.

The terrifying thing was that my internal world was still as complex and meaningful as it was before. All of the emotions I felt were real and legitimate. My cognition outside of lqnguage was intact, I could do math just fine and conceptualize and abstract problems.

In spite of this, I was unable to convey what I was feeling and thinking to the outside world. It felt like I was trapped inside my own body.

After recovering, my intuitive understanding of the link between language and intelligence was changed forever.

See the following for more info on Topamax/Topiramate language impairment: https://link.springer.com/article/10.1007/s10072-008-0906-5

Re: Telling GPT-4 you're scared or under pressure improves performance

#207

Post by intended »

williamcotton wrote: post #201
intended wrote: post #193

Earlier quoted context omitted.

This paper was published, and the distinct jump was found to be a measurement artifact. https://hai.stanford.edu/news/ais-ostensible-emergent-abilit...

Sweet, now complete and present a meta-analysis of all such papers about the topic and maybe I’ll find it more convincing than a cherry-picked publication that supports your preexisting position on the topic!

I cherry picked a Stanford article and professor… those Stanford liars! Curse them!

and silly me! You are absolutely right, research should have a meta analysis ready of papers that came out… 6 months ago.

Absolutely ridiculous.

I am sure you have read this paper, and have an army of meta analyses and counter claims for this nascent sub field.

Re: Telling GPT-4 you're scared or under pressure improves performance

#208

Post by RoyalHenOil »

qingcharles wrote: post #202
famouswaffles wrote: post #167

Earlier quoted context omitted.

>“You” are merely a passenger, observing your physical self’s actions and assuming ownership of them Oh boy. Now here's the rub. We know this is true at least sometimes. When you make a decision and explain it, it is often(always?) just a post-hoc rationalization. You don't actually know why you make a lot of the decision you do even if you dearly believe so.

This is very clear in split-brain patients where one side will make the decision and the other side will come up with an excellently reasoned explanation of the decision.

What's even weirder is that when the hemispheres are split apart, the two sides seem to both be able to understand and carry out instructions (and can even compete with each other to do them, e.g., the left hand and right hand may try to push each other out of the way), which suggests that there may be two wholly separate consciousnesses in each half of the brain.

Re: Telling GPT-4 you're scared or under pressure improves performance

#209

Post by RoyalHenOil »

mastercheif wrote: post #206
SamFold wrote: post #198

Earlier quoted context omitted.

One interesting point here is that humans can learn to do things without language at all. If you raised a human baby and never exposed it to any language it would still learn skills and behaviors. So while intelligence and reasoning in humans does still seem to be linked to language it’s not quite as simple as all knowledge and reasoning simply being encoded in language. Whereas (obviously) ChatGPT is completely base…

In college I tried a medication called Topamax (Topiramate) for migraine prevention. Topamax has a low-occurrence side effect of “language impairment” which I was particularly sensitive to. It was a terrifying experience, but it was also a valuable one as it changed the way I view intelligence. 10 days in my writing and speech skills had devolved to that of a primary school aged child. I tried to type a text message…

I do not experience an internal monologue (unless I make a conscious effort at it), and I also struggle to find the right words for things. For this, and for a few other reasons (like taking a longer time than usual to understand speech), I suspect I may have a mild language processing disorder.

But my inner world is still extremely rich. I have very complex thoughts, can solve complex puzzles, etc., all without thinking any words.

It sounds like your medication gave you a particularly intense experience of this.

Many people have historically downplayed the consciousness of animals and babies on account of their inability to understand language, but it seems language likely has nothing to do with consciousness.

Re: Telling GPT-4 you're scared or under pressure improves performance

#210

Post by suzzer99 »

Ldorigo wrote: post #203
suzzer99 wrote: post #15

Am I the only one who feels bad for asking ChatGPT a "dumb" question that I know I should know, or not saying thank you when it gives me an answer? No? I'm just a weirdo? Okay. I have to push back really hard against my proclivity to humanize it, to the point where I probably don't use it as much as I should, just because I don't want to deal with the psychic stress of reminding myself that it's not a living entity.

I mostly feel bad because I don't want to know the energy costs of having my dumb questions answered this way. I also feel really guilty when asking chatgpt for "easy" stuff that a regex could do (like normalizing whitespace in a paragraph and the likes). Really curious how many Watts/token gpt4 uses.

Regex is like half my questions to it. Hmm 15 minutes fighting with Javascript regex weirdness (wait do I have to backslash the backslash that backslashes the slash?), or 20 seconds with ChatGPT.
Post reply on HN