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

Ask HN: Do you still use search engines?

news.ycombinator.com

Post reply on HN
Print view View original

531–540 of 609 posts

Re: Ask HN: Do you still use search engines?

#531

Post by brookst »

octernion wrote: post #351
wavemode wrote: post #328

Earlier quoted context omitted.

Yes, I have. If I want something to read the page for me, then that's where LLMs come in. But what I'm talking about is when I want to read the page for myself. Waste of time to have to wait for an LLM to chew on it.

hah i use llms for that now too - "option-space 'link to lang'" and chat returns faster than the whole endeavor of opening google or putting stuff into the nav bar.

That’s my experience too. I don’t find Google or even Kagi faster for retrieving a link. All of the major LLMs can pull search results faster than I can through search websites, and then I’ve got a conversation where I can dive deeper if I want.

Re: Ask HN: Do you still use search engines?

#532

Post by wruza »

teruakohatu wrote: post #515
wruza wrote: post #453

Earlier quoted context omitted.

Is there a well-known benchmark for this? I don't feel that short vs long answers make any difference, but ofc feelings aren't what we can measure. Also, if that works, why doesn't copilot/cursor write lots of excessive code mixed with lots of prose only to distill it later?

> don't feel that short vs long answers make any difference The “thinking” models are really verbose output models that summarise the thinking at the end. These tend to outperform non-thinking models, but at a higher cost. Anthropic lets you see some/all of the thinking so you can see how the model arrived at the answer.

So if I replace "answer" with "summarize" that should work then?

Re: Ask HN: Do you still use search engines?

#533

Post by smallerize »

saltcured wrote: post #171
pdabbadabba wrote: post #107

Earlier quoted context omitted.

If what you say is literally true: yes, I think you probably are the only one!

Yeah, I need more coffee to decide for myself if double checking all sources is linear or exponential as it progresses to check the checks.

The sources will start to be redundant eventually. It's actually O(1) once you have looked at all the sources... that there are... in the world. Trivial!

Re: Ask HN: Do you still use search engines?

#534

Post by nkrisc »

nh23423fefe wrote: post #295
nkrisc wrote: post #285

Earlier quoted context omitted.

> it seems like youve internalized the query syntax of search engines If by syntax you mean "words related to what I want to find" then I suppose. > youre fine with erroneous results. Because AI is famously correct all the time. And there are no "erroneous" results in my method, only ones not closely enough related to what I wanted, so that's feedback my search terms should be refined. Any answer I get from AI I’m go…

no by query syntax i mean what people have called google fu. yeah seems the same to me. type some stuff, read the response, evaluate, loop or break. cool you hate ai and dont want to learn a new pattern. but you say i dont want to think which doesnt make sense. you are thinking just as hard.

> what people have called google fu

I don’t know what that means in any specific sense.

I don’t want to think about how to interact with the search mechanism. It’s the same problem with using voice assistants like Siri - I have to think about how to construct my query so that it will be parsed correctly.

With most search engines I can just type in disparate terms that might be related. The order doesn’t matter, the phrasing doesn’t matter, I don’t need to give it instructions on how to respond, etc.

Re: Ask HN: Do you still use search engines?

#536

Post by Shorel »

I still use search engines, and not ChatGPT or any LLM as my primary behaviour.

Of course, I have used Phind and other LLMs, and the results sometimes are useful, but in general the information they give back feels like a summary written for the “Explain Like I'm Five” crowd, it just gives me more questions than answers, and frustrates me more than it helps me.

Where LLMs excel is when I don't know the exact search term to use for some particular concept. I ask the LLM about something, it answers with the right terms I can use in a search engine to find what I want, then I use these terms instead of my own words, and what I want is in the search results, in the first page.

Re: Ask HN: Do you still use search engines?

#537

Post by Shorel »

RyanOD wrote: post #376
Modified3019 wrote: post #264

Earlier quoted context omitted.

IMO an example of a good use case for an LLM, which would be otherwise very hard to search for, is clarifying vague technical concepts. Just yesterday I was trying to remember the name of a vague concept I’d forgotten, with my overall question being: “Is there a technical term in biology for the equilibrium that occurs between plant species producing defensive toxins, and toxin resistance in the insect species that f…

Agreed. I'm increasingly using ChatGPT to research topics. In that way, I can refine my question, drill down, ask for alternatives, supply my own supplementary information, etc. I think of AI as an intelligent search engine / assistant and, outside of simple questions with one very specific answer, it just crushes search engines.

I don't see it as crushing the search engines.

I use the LLMs to find the right search terms, and that combination makes search engines much more useful.

LLM by themselves give me very superficial explanations that don't answer what I want, but they are also a great starting point that will eventually guide me to the answers.

Re: Ask HN: Do you still use search engines?

#538

Post by udev4096 »

Yes, why would I not? I, unlike you, do not intend to have a shallow knowledge on things I wanna know about. In a few years, it's going to get worse and no one would have deep expertise on anything (especially junior engineers) if they keep using LLMs. DDG is still far better than Google although I have started to see more ads on DDG searches which is quite annoying

Re: Ask HN: Do you still use search engines?

#539

Post by udev4096 »

agubelu wrote: post #298
exe34 wrote: post #275

Earlier quoted context omitted.

that sounds like a really great example of..... searching through vector embedding. I don't think the LLM part was technically necessary.

It's all about the UI/UX. Same as in Dropbox/rsync

No one who extensively uses dropbox uses their UI. It's mostly either the API or rclone. It's inefficient and a complete waste of time to use any UI at all

Re: Ask HN: Do you still use search engines?

#540

Post by dumbfounder »

wavemode wrote: post #69

Search is primarily a portal - you know a particular resource exists, you just don't know its exact URL. You hear about this new programming language called "Frob", and you assume it must have a website. So you google "Frob language". You hear that there was a plane crash in DC, and assume (CNN/AP/your_favorite_news_site) has almost certainly written an article about it. You google "DC plane crash." LLMs aren't ever…

Yes they will. Why do you think they won’t? They certainly can. You just use RAG to look up the latest news based on the keywords you are using. You can use search on the back end and never surface a list of results unless the LLM decides that is a good idea. It curates that the reusits for you. Or gives you the singular site you need with context. That is better for most searches.
Post reply on HN