Earlier quoted context omitted.
by "eventually" they mean "at this rate, mid-next year, if not already"
No - they are all competing for eyeballs right now. They need to solidify their positions in terms of user habits and behaviors before they start pissing people off.
Ask HN: Do you still use search engines?
511–520 of 609 posts
Re: Ask HN: Do you still use search engines?
#512I still prefer tranditional search engines over LLMs but I admit, its results feels worse than it has traditionally. I don't like LLMs for two reasons: * I can't really get a feel for the veracity of the information without double checking it. A lot of context I get from just reading results from a traditional search engine is lost when I get an answer from a LLM. I find it somewhat uncomfortable to just accept the a…
This is my main reason for not using LLMs as a replacement for search. I want an accurate answer. I quote often search for legal or regulatory issues, health, scientific issues, specific facts about lots of things. i want authoritative sources.
Re: Ask HN: Do you still use search engines?
#513Re: Ask HN: Do you still use search engines?
#514Re: Ask HN: Do you still use search engines?
#515Earlier quoted context omitted.
Yeah but it kind of kneecaps the model. They need tokens to "think". It's better to have them create a long response then distill it down later.
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?
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.
Re: Ask HN: Do you still use search engines?
#516Re: Ask HN: Do you still use search engines?
#5171. *Browsing*
This can be completely avoided. Here is a thing you can do on firefox with some tweaks in order to achieve no-search browsing
- Remove search suggestions in (about:preferences#search)
- Use the [History AutoDelete](https://addons.mozilla.org/en-US/firefox/addon/history-autod...) addon to remove searches from your History. This will avoid searches from your history to pollute the results
- Go to (about:config) and set `browser.urlbar.resultMenu.keyboardAccessible` to `false`
Now when you Ctrl + L into the tab, you will get results from your history, bookmarks and even open tabs. And the results are only a few Tab presses away, no need to move your hands off the keyboard.
If you don't like the results and want to launch a search anyways, well just press Enter instead and it will launch a search with the default search engine. A cool trick is to type % + space in the awesome bar to move around opened tabs/ You can also specifically look into bookmarks with * and history with ^
P.S : Ctrl + L, Ctrl + T, Ctrl + W and Ctrl + Alt + T are your best friends.
P.P.S: Now you can also learn more about custom search engines : https://askubuntu.com/a/1534489
2. *Quick answer* on a topic. This is the second most common use case and what Google has been trying to optimize for for a long time. Say you want to know how many people are there in Nepal or what is the actual percentage blue-eyed people in Germany. This is where llm shine I think but to be fair Google is just as good for this job.
3. *Finding resources* to work with. This one is a bit on the way out because, it's what people who want to understand want but we probably are few. This is valuable because those resources do not just give an answer but also provide the rationale/context/sources for the answer. But.
On the one hand, most people just want the answer, and most people can be you if, even though you deem yourself a curious person, you don't have the time right now to actually make the effort to understand. On the other hand, llms can craft tutorials and break down subjects for you which turn those resources much less valuable. I kind of feel like the writing is on the wall and the future for this use case is for "curating" search engines that will give you the best resources and won't be afraid to tell you "Nothing of value turned up" instead of giving you trash. Curious to hear your thougts about that.
Re: Ask HN: Do you still use search engines?
#518It is easy to filter them when you working with familiar domain, but trying to learn something completely new - it is better to ask DeepSeek for a summary, and then decide what to explore.