One of Google Search’s oldest and best-known features, cache links, are being retired. Best known by the “Cached” button, those are a snapshot of a web page the last time Google indexed it. However, according to Google, they’re no longer required.
“It was meant for helping people access pages when way back, you often couldn’t depend on a page loading,” Google’s Danny Sullivan wrote. “These days, things have greatly improved. So, it was decided to retire it.”
How has no one worked on a new search engine over the last decade or so where Google has been on a clear decline in its flagship product!
I know of the likes of DDG, and Bing has worked hard to catch up, but I’m genuinely surprised that a startup hasn’t risen to find a novel way of attacking reliable web search. Some will say it’s a “solved problem”, but I’d argue that it was, but no longer.
A web search engine that crawls and searches historic versions of a web page could be an incredibly useful resource. If someone can also find a novel way to rank and crawl web applications or to find ways to “open” the closed web, it could pair with web search to be a genuine Google killer.
My favorite case in point being Google Maps.
There’s a lot of startups trying to make better search engines. Brave for example is one of them. There’s even one Lemmy user, but I forget what the name of theirs is.
But it’s borderline impossible. In the old days, Google used webscrapers and key word search. When people started uploading the whole dictionary in white text on their pages, Google added some antispam and context logic. When that got beat, they handled web credibility by the number of “inlinks” from other websites. Then SEO came out to beat link farmers, and you know the rest from there.
An indexable version of Archive.org is feasible, borderline trivial with ElasticSearch, but the problem is who wants that? Sure you want I may, but no one else cares. Also, let’s say you want to search up something specific - each page could be indexed, with slight differences, thousands of times. Which one will you pick? Maybe you’ll want to set your “search date” to a specific year? Well guess what, Google has that feature as well.
Cached versions can sometimes get around a paywall when a site gives Google access but charges users.
Archive.is them
The next revolutionary search engine will be an AI that understands you. Like what a librarian is… Not just ads served.
i don’t need a search engine that understand me i need a search engine that finds sites and pages based on a string of text i provide it
we should be calling the future piss the way it’s going down the toilet
Well, at the least, you need something to filter out the shit trying to game seo. To me it seems that AI is the easiest approach.
I recommend Kagi. Bought a family plan and it feels like I’ve gone back to 2016 when the search engines weren’t a dumpster fire.
Second kagi. I’m just on the personal plan, but can confirm it’s fire
Bing’s copilot is genuinely pretty good, the AI answer is often pretty accurate and the way it’s able to weave links into its answer is handy. I find it way more useful than Google search these days and I’m pretty much just using it on principle as Google is just pissing me off with killing their services, a few of which I’ve used.
I don’t think Microsoft is some saint but copilot is just a good product.
Yes, that would be a Google killer. If you somehow find the money to provide it for free.
Finding a novel way of searching is one thing. Finding a novel way of financing the whole endeavor (and not going the exact route Google is) is another.