Follow Slashdot blog updates by subscribing to our blog RSS feed

 



Forgot your password?
typodupeerror
The Internet Google Social Networks

Reddit Wants To Be a Search Engine Now (theverge.com) 41

Reddit wants to become a full-fledged search engine, leveraging its vast repository of human-generated content and expanding its AI-powered Reddit Answers tool. In its latest note (PDF) to investors, CEO Steve Huffman says the company is "concentrating our resources on the areas that will drive results for our most pressing needs," including "making Reddit a go-to search engine." The Verge reports: Huffman says that "every week, hundreds of millions of people come to Reddit looking for advice, and we're turning more of that intent into active users of Reddit's native search." Reddit's core search has more than 70 million weekly active unique users -- Reddit overall averages 416.4 million weekly active unique users -- and Reddit Answers, the platform's AI search tool that it launched in December, has 6 million weekly users, up from 1 million weekly users in the first quarter of this year. To continue to build out search, Reddit is "expanding Reddit Answers globally, integrating it more deeply into the core search experience, and making search a central feature across Reddit," Huffman says.

Reddit Wants To Be a Search Engine Now

Comments Filter:
  • April 1st again? (Score:5, Informative)

    by siege72 ( 1795922 ) on Friday August 01, 2025 @06:50AM (#65559668)

    Reddit's internal search... infamously bad.

    It's more useful to use Google with the "site:" parameter than stay within Reddit to search.

    • by mccalli ( 323026 )
      It's great! I mean, the syntax is a little unusual in that it seems to start with the URL google.com and the protocol isn't http or https but "site:reddit.com"...other than that, the search is great!
  • by Tough Love ( 215404 ) on Friday August 01, 2025 @06:54AM (#65559670)

    Does reddit want to be a shitty search engine like wikipedia? Not that I don't get a whole lot of use out of wikipedia's search engine - in fact the vast majority of my searches go through wikipedia's engine and not google's. Suck it google, you don't get to spy on the vast majority of my searches. But this works because I feed wikipedia really softball searches. Send it the sort of search you regularly throw at google and it just barfs.

    Wikipedia has not got anything remotely resembling Pagerank and there doesn't appear to be any agenda to create such a thing. You can't just copy google's (now out of patent) algorithms because the situation is different. You don't have zillions of netizens linking things. You've got something much more regimented happening, and in the past every attempt to extend Pagerank to non-internet situations has met with failure. Including google's. Did you know they used to offer corporate search appliances that they hyped as like google but private and entirely inside your own network? Abject failure. Nobody making links means the search engine just sucks way too much, like old school alta vista or hotbot. Tons of results with zero ranking. Pull your hair out. Slit your wrists. Doesn't matter, it won't work any better no matter what you do. After years of throwing money at it they finally canned the whole department.

    OK, is that what reddit wants? If that is not in fact what reddit wants then they are going to have to swallow their pride and bring in some people who can invent what they want, and implement it reliably. Not going to be cheap. I would suggest playing the open source card on this one, otherwise... welcome to the shitty internal search club.

    • by DarkOx ( 621550 )

      I remember that; there was a ton of hype around Google's internal search thing. I never saw it do anything better than basic full text search already present in SQL/Sharepoint/Exchange

      The problem is software or hardware it is 'how' do you do it in a useful way. Page range for all the fancy academic writing about it boils down to don't actually do it, let humans do it, and pick up on the fact the content is interesting by how much linking to it there is/was.

      One kind stupid but maybe not stupid approach I c

      • I never saw it do anything better than basic full text search already present in SQL/Sharepoint/Exchange

        Don't need any more proof than that to know those self-described smart people aren't all that smart. Not that other convincing demonstrations are in anything remotely resembling short supply.

      • BTW I like your line of thinking. Actual smart person as opposed to the vast majority of google posers.

    • by HiThere ( 15173 )

      AltaVista gave me more accurate results than Google ever has, much less recently. But you did need to use reasonable logic, and "near" was an important addition.

      Google was easier to use. That was their sole advantage.

      • Altavista was goddamn terrible. It wasnt just the ease of use, its that pagerank was *clearly* a superior algorithm than the weird fuzzy thing altavista did that sort of ensured you'd *maybe* find a useful result a couple of pages in if you where really careful with your search booleans. Google was ......... magic......... in the beginning.

    • PageRank doesn't really work on the Internet any more either - how often does anyone link to the things they're talking about? At best you might have some references at the bottom of a page, but most of the time, sites don't really have hyperlinks in the text. As such, a lot of the context for pagerank is lost, and all you have is "number of links", which is essentially link farming.

      That said, Google do still have some notion of "reputation" for web sites - higher rep means higher results. I suspect that's

      • PageRank started breaking after Google acquired DoubleClick in 2007/8. If you survey search savvy people about when Google search results started getting noticeably worse, they'll usually point at 2010-2015 - basically as DoubleClick's influence internally took hold. Google search feeding absolute garbage AI generated content heavily was natural pivot as Google has their own vested interest in the success of current generative systems. This was a rising tides raise all ships decision from the top. And ste
  • Is it the Apple Car of Gated Communities?
  • by DeplorableCodeMonkey ( 4828467 ) on Friday August 01, 2025 @07:58AM (#65559716)

    The general progression toward using AI to directly answer search queries is going to be a huge problem for content producers, and I don't think the AI firms or Google appreciate the extent of the coming fight.

    The argument used to be "Google provides you value by making your content discoverable." Now, AI companies consume content and provide their own takes on it without even referring people back to the original content. Result: no revenue at all from that activity for the content provider and a tiny loss from having to serve up the content for ingestion.

    Long term, expect a growing market for paywall solutions for content providers and more companies like Reddit responding to Google, OpenAI and Anthropic as though they are competitors which they really are now.

    On the flip side, this could lead to genuine innovation in fintech to make it painless to pay for content to support good providers which could create a lot of opportunity for the little guy in the content creation markets.

    • On the flip side, this could lead to genuine innovation in fintech to make it painless to pay for content to support good providers which could create a lot of opportunity for the little guy in the content creation markets.

      Anything that doesn't take the web back to millions of tiny pages made by random people then it sucks shit.

    • The general progression toward using AI to directly answer search queries is going to be a huge problem for content producers, and I don't think the AI firms or Google appreciate the extent of the coming fight.

      Yup, and it already is. Web searches are already useless, and AI is turning out to be eating into those web searches. Definitely an increasing problem.

      Example - I use DDG for searches.

      In a recent search for someone who drove off a bridge maybe 12 years ago because their GPS "told" them to, DDG (and Google since I tried them as well) returned page after page of the same results. News stories about a guy who drove off a bridge in 2023. Just from every news provider. Useless.

      I knew it was a man and wife -

      • by DarkOx ( 621550 )

        You had an answer...Are you sure it was correct?

        • You had an answer...Are you sure it was correct?

          Yes, I took the answer and cross referenced it with names and dates it provided. In this case, the AI search was 100 percent correct.

          It is true that AI generated answers can and are wrong at times. Sometimes hilariously wrong. It isn't perfect.

          Then again, many pages of the same incorrect story even when I provided the details I knew, is not only not correct for my search, it is 100 percent useless. There have been complaints that web search is down. It is, and it deserves to be, because it is a waste

      • DDG: "Guy and wife drove off a bridge 12 years ago" https://duckduckgo.com/?q=guy+... [duckduckgo.com] "Search assist (AI) - generate an answer" -> "sorry, no relevant information was found".
        However, the correct answer from CNET is third result from DDG https://www.cnet.com/culture/m... [cnet.com] The anecdote is mentioned in Wikipedia https://en.wikipedia.org/wiki/... [wikipedia.org]

        • DDG: "Guy and wife drove off a bridge 12 years ago" https://duckduckgo.com/?q=guy+... [duckduckgo.com] "Search assist (AI) - generate an answer" -> "sorry, no relevant information was found".

          My guess is that I didn't use those exact words and was reconstructing it from memory. Probably used "man and wife drove off a bridge GPS" The AI results:

          "Overview of GPS-Related Bridge Accidents

          Recent Incidents

          Two notable incidents highlight the dangers of relying on GPS navigation when driving near unsafe bridges:

          Philip Paxson Incident (2022): Philip Paxson drove off a collapsed bridge in Hickory, North Carolina, while following Google Maps directions. The bridge had been unmarked and unbarricaded si

    • by Jerrry ( 43027 )

      I think the rise of AI, along with Reddit's often harsh and silly moderation decisions, is going to result in Reddit becoming irrelevant just like Stack Overflow.

  • by Sean Clifford ( 322444 ) on Friday August 01, 2025 @08:27AM (#65559758) Journal

    If the goal of Reddit is the suicide of its residents, then yes, by all means, cripple Reddit with AI.

  • That's ironic, because Google has apparently decided it doesn't want to be a search engine anymore.
  • Awesome monetization of half-baked opinions and less-well-baked Domino's pizza. JFC, how low the bar is set theses days.

  • by jpellino ( 202698 ) on Friday August 01, 2025 @08:41AM (#65559786)

    He means money.
    A search engine now makes money while almost inadvertently supplying useful search results.
    See google results for the past year, shopping is now the top tier of results.
    Sponsored results too. Gotta ad-incentivize those results.
    Since COVID they have also convinced students that they are an operating system.
    Used to be people gave up on a search 3 or 4 clicks deep. I believe that number now approaches 1.
    Sorry, usually only this cynical *before* breakfast.

  • ...more shitty opinions on things we didn't search for.

  • They can want. That doesn't mean that it will happen.

    COnsidering how famously atrocious Reddit's search function is, there is zero change of them being a serious search engine.

    Google is being toppled by AI(ChatGPT/Copilot), but Reddit ranks as perhaps the worst search engine ever.

  • Considering that more than half of Reddit is either false or fake, this is going to end really well. I give it 5 years before the whole AI bullshit collapses into a "stupid people will believe anything" meme. This no different from all the execs who "cloudified" all their system to save money and got huge bonusus followed by all the new execs who are starting there in-sourcing project to take everything out of the "cloud" and back into personal datacenters, you guessed it, because the "cloud" doesn't work

    • by Ksevio ( 865461 )

      But now people can save a step and go to reddit search to be told to eat paste rather than needing to type site:reddit.com in their Google query

  • by ledow ( 319597 )

    I like Reddit.

    I've literally never even clicked its AI thing in the app or on the website, and I only discovered it was even there a few days ago. My eyes literally gloss over anything that's not the stuff I use regularly (I don't care about stickers, coin, trends, or all the other nonsense either).

    Stick to what you know. I don't frequent a barber shop because I hope one day it'll also sell me travel insurnace. I can't see myself using Reddit as a search engine unless it quite literally turns out to be t

    • I like Reddit.

      I've literally never even clicked its AI thing in the app or on the website, and I only discovered it was even there a few days ago. My eyes literally gloss over anything that's not the stuff I use regularly (I don't care about stickers, coin, trends, or all the other nonsense either).

      Stick to what you know. I don't frequent a barber shop because I hope one day it'll also sell me travel insurnace. I can't see myself using Reddit as a search engine unless it quite literally turns out to be the

  • It's the CIIIIIIRCLE of SLOP!

    Now LLMs trained on Reddit slop are going to trawl through Reddit slop and produce more slop.... on which the next generation will be trained.

    BRILLANT

  • I'm betting Bing will be a grown-up search engine long before Reddit. It's *not* easy.

  • I've been consuming Reddit on a daily basis for many years. For pure entertainment.

    Without any shade of doubt, the quality of recent content has degraded towards a vast majority of slop content and comments. Mayhaps interesting human content is just depressed for my anon visit, who knows.

    But for sure, using Reddit content since 2022 as a source for search or AI training will indeed be feeding the AI dog its own tail.

Computers can figure out all kinds of problems, except the things in the world that just don't add up.

Working...