Hacker News Reader

Comment by puppycodes | Hacker News Reader

Here's one simple reason:

I have a very specific esoteric question like: "What material is both electrically conductive and good at blocking sound?" I could type this into google and sift through the titles and short descriptions of websites and eventually maybe find an answer, or I can put the question to the LLM and instantly get an answer that I can then research further to confirm.

This is significantly faster, more informative, more efficient, and a rewarding experience.

As others have said, its a tool. A tool is as good as how you use it. If you expect to build a house by yelling at your tools I wouldn't be bullish either.

View on HN

Until they pull the rug and make you read through 10 paragraphs of ads first before getting to your answer.

I mean, the first link I got when I pasted that in is probably the Stack Exchange thread you would use to research further, along with other sources, which do seem relevant to the query.

I don't see how an LLM is significantly faster or more informative, since you still have to do the legwork to validate the answer. I guess if you're google-phobic (which a lot of people seem to be, especially on HN) then I can see how it's more rewarding to put it off until later in the process.

you can boil it down to this, whats easier for most people? looking through websites and search engine results for answers or speaking in plain language? The answer is pretty obvious.

The validity of the answers is not 1:1 with its potential profitability.

Like James Baldwin said "people love answers, but hate questions."

getting an answer faster is exponentially better than getting the more precise, more right, more nuanced answer for most people every time. Doing the due dilligence is smart but its also after the fact.

Its just an example and you can "well actually" until the cows come home, but your missing the point. I'm sure there are things you have found hard to google. Also you are most likely (as your commenting on hacker news) not a good representation of the majority of the world.

idk if you have noticed, but google is clearly using LLM technology in conjunction with its search results, so the assumption they are just using traditional tech and not LLM's to inform or modify its result set I think is unlikely.

> I'm sure there are things you have found hard to google.

There are, but I haven’t found LLMs to be useful for them either.

Today, I had the question of “where can I buy firewood within a ten minute drive from my house, and what’s the cost at each place?”

There’d no real way to get that info without going for a drive, or calling every potential location to ask.

The first stackexchange link I see answers the question of thermal conductivity, not electrical. Google is convinced I didn’t actually mean electrical. Forcing it to include electrical brings up nothing of use.

The Google AI summary suggests MLV which is wrong.

ChatGPT suggests using copper which is also wrong.

I call bullshit on the entire affair.

The answer I got from ChatGPT was: ...........

A material that is both electrically conductive and good at blocking sound is:

Lead (Pb) • Electrical conductivity: Lead is a metal, so it conducts electricity, although it’s not the most conductive (lower than copper or silver).

• Sound blocking: Lead is excellent at blocking sound due to its high density and mass, which help attenuate airborne sound effectively.

Other options depending on application:

Composite materials:

• Metal-rubber composites or metal-polymer composites can be engineered to conduct electricity (via embedded conductive metal layers or fillers) and block sound (due to the damping properties of the polymer/rubber layer).

Graphene or carbon-filled rubber:

• Electrically conductive due to graphene/carbon content.

• Sound damping from rubber base.

• Used in some specialized industrial or automotive applications.

Let me know if you need it optimized for a specific use case (e.g., lightweight, flexible, non-toxic).

...........

This took me less than 10 seconds.

Pretty damn good if you ask me.

You used the same search string and got a completely different answer?

That’s… weird. Definitely doesn’t inspire confidence.

I have a prompt personalization that says im a scientist / engineer. Perhaps thats why it gave me a better answer. If you consider the multitude of contexts you could ask this question it makes sense to give it a little personal background.

Reply on news.ycombinator.com