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

Writing is thinking

nature.com

Post reply on HN
Print view View original

61–70 of 135 posts

Re: Writing is thinking

#61

Post by wenc »

nicolapede wrote: post #43
spyckie2 wrote: post #31

Earlier quoted context omitted.

I really think the effects of LLMs on thinking is the exact same as a calculator. It shortcuts some forms of thinking to open up other forms of thinking. My thinking has increased with the use of LLMs, not decreased, most likely because LLMs take the edge off of grind work like reading a lot of noise to capture the 1% signal, formulating accurate statements for abstract ideas, and bringing together various domains th…

>> But allowing calculators WILL make those who like math reach much deeper into the field than without. Have you ever run into any mathematician that praised the calculator for his/her career? I’d be really curious to read about that.

Pure math people probably don’t reach for calculators. But engineers do all the time. Back of the envelope guesstimating is bread and butter.

The modern equivalent of a calculator is Excel.

Re: Writing is thinking

#62

Post by dist-epoch »

pcrh wrote: post #56

I am quite puzzled how an LLM could even start "write" a scientific paper. Say you start with a set of findings, for example, western blots, data from a transgenic mouse engineered for the relevant gene, and some single cell sequencing data. Your manuscript describes the identification of a novel protein, editing the gene in a mouse and showing what pathways are affected in the mouse. What material would you give the…

> How would the LLM "know" which of these novel findings were in any way meaningful

Given that they are trained on all of arXiv, ..., it's much more likely they are aware of all public relevant papers than your average researcher.

Re: Writing is thinking

#63

Post by dboreham »

pcrh wrote: post #56

I am quite puzzled how an LLM could even start "write" a scientific paper. Say you start with a set of findings, for example, western blots, data from a transgenic mouse engineered for the relevant gene, and some single cell sequencing data. Your manuscript describes the identification of a novel protein, editing the gene in a mouse and showing what pathways are affected in the mouse. What material would you give the…

The LLM can make a plan or outline first, which is also writing.

Re: Writing is thinking

#64

Post by kelseyfrog »

Writing needs a conceptual split analogous to the split between math and calculating.

Just as a calculating can be implemented on a computer which has low cognitive abilities but high algorithmic and procedural abilities, we need to extract out the word-smithing capabilities from writing separate from the thinking portion. Our lack of distinction in terms reflects a muddled conceptual framework.

LLMs are excellent wordsmiths completely divorced from the concept of thinking. They break the correlative assumption - that excellent writing is corresponds with excellent thinking. Until now, we've been able to discern poor idea because they have a certain aesthetic, think conspiracy rants in docx saying something about a theory of everything based on vibrations. But that no longer holds. We have decent enough word-smithing coupled with a deficit of thinking. Unfortunately this breaks our heuristics with consequences ranging from polluting our online commons to folks end up believing nonsense like ChatGPT named itself Nova and they are a torchbearer for spiritual gobbledygook.

My point is that we're in the process of untangling these two and as a result, we're likely to see confusion and maybe even persistent misunderstanding until this distinction becomes a more common part of how we talk about and evaluate written work. They're living in an AGI-world and we're just..not.

Re: Writing is thinking

#66

Post by zkmon »

Whenever I need to do some hard thinking and things are not clear, I fire up my sublime text and write down the context in the simplest terms and short lines (only few words per line). While doing this, I will be absolutely rude to myself asking extremely basic and direct questions to bring out the real context, real goals and real path. It's like answering an under-world boss. No bullshit, no pretense, no regard to any norms, no impressing someone. Then the whole thing falls into a meaningful structure. I leave it when it produces some immediate action items.

Re: Writing is thinking

#67

Post by lukebechtel »

zug_zug wrote: post #2

I see a lot of people say "writing is so important," and I think what they mean is "I feel really smart/good when I write." And I think what they are experiencing is that they've been assembling ideas in their heads for weeks, and only when it's all come together are they ready synthesize that information at a higher level, and they mistake this synthesis for the writing itself (rather than the writing being a sympto…

it depends how you write.

often when I write an idea down, my "inner critic" process gets more activated upon seeing the textual representation.

thus I find gaps and flaws more easily.

not true for all domains, but many.

Re: Writing is thinking

#68

Post by pcrh »

dboreham wrote: post #63
pcrh wrote: post #56

I am quite puzzled how an LLM could even start "write" a scientific paper. Say you start with a set of findings, for example, western blots, data from a transgenic mouse engineered for the relevant gene, and some single cell sequencing data. Your manuscript describes the identification of a novel protein, editing the gene in a mouse and showing what pathways are affected in the mouse. What material would you give the…

The LLM can make a plan or outline first, which is also writing.

Any researcher already has this in their head long before any writing takes place.

Re: Writing is thinking

#70

Post by pcrh »

dist-epoch wrote: post #62
pcrh wrote: post #56

I am quite puzzled how an LLM could even start "write" a scientific paper. Say you start with a set of findings, for example, western blots, data from a transgenic mouse engineered for the relevant gene, and some single cell sequencing data. Your manuscript describes the identification of a novel protein, editing the gene in a mouse and showing what pathways are affected in the mouse. What material would you give the…

> How would the LLM "know" which of these novel findings were in any way meaningful Given that they are trained on all of arXiv, ..., it's much more likely they are aware of all public relevant papers than your average researcher.

A researcher on any particular topic is not supposed to be an "average" researcher, but already deeply familiar with their subject.
Post reply on HN