Ask HN: Do you still use search engines?
Today, I noticed that my behavior has shifted over the past few months. Right now, I exclusively use ChatGPT for any kind of search or question.
Using Google now feels completely lackluster in comparison.
I've noticed the same thing happening in my circle of friends as well—and they don’t even have a technical background.
How about you?
608 comments
[ 5.7 ms ] story [ 324 ms ] threadThe more times goes by, the more I use both ChatGPT and Claude to search (at the same time, to cross-check the results) with Kagi used to either check the results when I know strictly nothing of the subject or for specific searches (restaurants, movie showings…).
I’ve almost completely stopped using Google.
If you go for the highest tier subscription on kagi, you get https://kagi.com/assistant which gives you a huge swath of AI models to handle your searching.
I think this also stems from a new design paradigm emerging in the search domain of tech. The content results and conversational answers are merging – be it Google or your Algolia search within your documentation, a hybrid model is on the rise.
I'm still using Google for searches on Reddit these days because Reddit's own search engine is terrible.
I feel like the google search will become obsolete in a short time and they have to make big changes to their UX and search engine.
Although I guess most of its user base are still relying on the old ways so changing it right now has huge impacts on older users.
For instance I wanted help cooking Coq au vin yesterday. I’ve cooked it before but I couldn’t remember what temperature to set the oven to. I read about five recipes (which were all wildly different) and choose the one that best suited the ingredients and quantities I was already using.
I asked chat gpt for a coq au vin recipe, and I’ll just say I won’t be opening a restaurant using ChatGPT as my sous chef anytime soon.
I can only really validate the generated response when it's code. Usually on other stuff, I trust and read the response which is not good I guess.
Hope you were satisfied with the food at the end :)
The other day I was also searching for something dumb: how to hammer a nail into concrete.
Google will find me instructions for a hammer-drill... no I just have a regular hammer. There's a link from wikiHow, which is okay, but I feel like it hallucinates as much as AI. Actually I just opened the link and the first instruction involves a hammer drill too. The second one is what I wanted, more wordy than ChatGPT.
Google then shows YouTube which has a 6 minute video. Then reddit which has bad advice half the time. I'm an idiot searching for how to hammer nails into a wall. I do not have the skill level to know when it's BS. Reddit makes me think I need a hammer drill and a fastener. Quora is next and it's even worse. It says concrete nails bend when hit, which even I know is false. It also convinces me that I need safety equipment to hit a nail with a hammer.
I just want a checklist to know that I'm not forgetting anything. ChatGPT gives me an accurate 5-step plan and it went perfectly.
Yes, I still use search engines and almost always find what I need in long form if I can’t figure it out on my own.
If it is more of an open ended question that I am not sure there'll be a page with an answer for, I am more likely to use ChatGPT/Claude.
A query in a regular search engine can at best perform like an LLM-based provider like Perplexity for simple queries.
If you have to click or browse several results forget it, makes no sense not to use an LLM that provides sources.
I just searched for "What is inherit_errexit?" at Perplexity. Eight sources were provided and none of them were the most authoritative source, which is this page in the Bash manual:
https://www.gnu.org/software/bash/manual/html_node/The-Shopt...
Whereas, when I searched for "inherit_errexit" using Google Search, the above page was the sixth result. And when I searched for "inherit_errexit" using DuckDuckGo, the above page was the third result.
I continue to believe that LLMs are favored by people who don't care about developing an understanding of subjects based on the most authoritative source material. These are people who don't read science journals, they don't read technical specifications, they don't read man pages, and they don't read a program's source code before installing the program. These are people who prioritize convenience above all else.
This makes a lot of sense to me. As a young guy in the 90's I was told that some day "everyone will be fluent in computers" and 25 years later it's just not true. 95% of my peers never developed their fluency, and my kids even less so. The same will hold try for AI, it will be what smartphones were to PCs: A dumbed down interface for people who want to USE tech not understand it.
Or I can just go to DDG/Google, and be done with it. No need to pre-load my "search engine" with context to get results.
Alt + Tab > Ctrl + T > Type > Enter > PgDn > Click > PgDn > Alt + Left > Click > PgDn > Alt + Left > Click > PgDn > Alt + Tab > [Another 45-60 minutes coding] > GOTO Start
With these keybinds (plus clicking mouse, yuck) I can read Nx sources of information around a topic.
I'm always looking to read around the topic. I don't stop at the first result. I always want to read multiple sources to (a) confirm that's the standard approach (b) if not, are there other approaches that might be suitable (c) is there anything else that I'm not aware of yet. I don't want the first answer. I want all the answers, then I want to make my own choices about what fits with the codebase that I am writing or the problem domain that I'm working in.
Due to muscle memory, the first four/five steps i can do in like one or two seconds. Sometimes less.
Switching to the browser puts my brain into "absorb new information" mode, which is a different skill to "do what IDE tells me to do". Because, as a software engineer, my job is to learn about the problem domain and come up with appropriate solutions given known constraints -- not to blindly write whatever code I'm first exposed to by my IDE. I don't work in an "IDE context". I work in a "solving problems with software context".
==
So I agree with the GP. A lot of posts I see about people saying "why not just use LLM" seem to be driven by a motivation for convenience. Or, more accurately, unconsidered/blind laziness.
It's okay to be lazy. But be smart lazy. Think and work hard about how to be lazy effectively.
[0]: not that I write blog post articles anyway, it's just a fantasy day dream thing that's been running through my head
They can be very useful, especially when looking for something closely adjacent to a popular topic, but you got to check carefully what they say.
Websites have all kinds of extra context and links to other stuff on them. If I want to learn/discover stuff then they are still the best place to go.
For simple informational questions, all of that extra context is noise; asking gpt "what's the name of that cpp function that does xyz" is much faster than having to skim over several search results, click one, wait for 100 JavaScript libraries to load, click no on a cookies popup and then actually read the page to find the information only to realise the post is 15 years old and no longer relevant.
There are times where I know exactly what website to go to and where information is on that site and so I prefer that over AI. DDGs bangs are excellent for this: "!cpp std::string" and you are there.
Then there's the verifiability thing. Most information I am searching for is code which is trivial to verify: sometimes AI hallucinates a function but the compiler immediately tells me this and the end result is I've wasted 30 seconds which is more than offset by the time saved not scrolling through search.
Examples of things that aren't easy to verify: when's this deprecated function going to be removed, how mature is tool xyz.
Of course, there's also questions about things that happened after the AI's knowledge cutoff date. I know there are some that can access the internet now but I don't think any are free
On the other hand every time I've used language models to find information I've gotten back generic or incorrect text + "sources" that have nothing to do with my query.
However, it still blinds you to the larger picture. Providing supporting sources is all well and good, but doesn't help you with the larger view. I want the larger view.
I use perplexity pro + Claude a lot as well. Maybe too much but mostly for coding and conversations about technical topics.
It really depends on intent.
I have noticed that I’ve started reading a lot more. Lots of technical books in the iPad based on what I’m interested in at the moment.
For me, searches fall into one of three categories, none of which are a good fit for LLMs:
1. A single business, location, object, or concept (I really just want the Google Maps or Wikipedia page, and I'm too lazy to go straight to the site). For these queries, LLMs are either overkill or outdated.
2. Product reviews, setup instructions, and other real-world blog posts. LLMs want to summarize these, and I don't want that.
3. Really specific knowledge in a limited domain ("2017 Kia Sedona automatic sliding door motor replacement steps," "Can I exit a Queue-Triggered Azure Function without removing it from the queue?"). In these cases, the LLMs are so prone to hallucination that I can't trust them.
The AI word salad summaries for each individual page have no toggle (unless you count !g).
Searching 'octatrack' has the wikipedia page on the top right as a summary box thing. no ai word salad for me :shrug:
The answer I'm seeking is not always on reddit itself, but google limited to reddit is far more likely to give me quality starting links than google unbound is.
At most I use AI now to speed up my research phase dramatically. AI is also pretty good at showing what is in the ballpark for more popular tools.
However I am missing forum style communities more and more, sometimes I don't want the correct answer, I want to know what someone that has been in the trenches for 10 years has to say, for my day job I can just make a phone call but for hobbies, side projects etc I don't have the contacts built up and I don't always have local interest groups that I can tap for knowledge.
If I need something more complex like programming, talk therapy, or finding new music then I’ll hop on over to Chat.