this post was submitted on 09 Dec 2024
-55 points (21.1% liked)

Technology

60106 readers
1923 users here now

This is a most excellent place for technology news and articles.


Our Rules


  1. Follow the lemmy.world rules.
  2. Only tech related content.
  3. Be excellent to each another!
  4. Mod approved content bots can post up to 10 articles per day.
  5. Threads asking for personal tech support may be deleted.
  6. Politics threads may be removed.
  7. No memes allowed as posts, OK to post as comments.
  8. Only approved bots from the list below, to ask if your bot can be added please contact us.
  9. Check for duplicates before posting, duplicates may be removed

Approved Bots


founded 2 years ago
MODERATORS
 

Reddit Answers uses AI to answer your questions with summarized information from across the platform.

An image showing the Reddit logo on a red and white background

Reddit is launching a new AI-powered search tool called Reddit Answers to help you more easily find information on the platform. The tool, similar to other AI search products, responds to queries by generating well-formatted responses and showing links to its sources.

But the distinction with Reddit Answers is that it sources things directly from Reddit, which means it could be a way to skip Google and get information directly from the source. Reddit has already been cracking down on where you can search for things from Reddit, and Google is the only major search engine that shows recent Reddit results. But Reddit probably would prefer if you did searches right on its platform, and Reddit Answers might prove to be a good way to do so.

Reddit Answers will initially roll out to “a limited number of users” in the US and in English, and it will be available on the web (not on Old Reddit) and iOS. (“We’re working on Android as we speak,” Serkan Piantino, Reddit’s VP of product, said in an interview.) Reddit plans to bring the tool to more languages and locations “in the future,” according to a blog post.

Reddit gave me access to a test version of Reddit Answers, and while I haven’t been able to mess around with it very much, I’ve liked what I’ve seen. There’s a big box to ask a query, but the initial Reddit Answers page also floats a bunch of suggested searches you can click on, like “favorite Nintendo character of all time,” “best mystery novels of 2025,” and “tips for flying with a baby for the first time.”

I clicked that last one, and the tool quickly pulled together a list and some bulleted suggestions with hyperlinks on the text and an arrow to click to see the source of the information. When you click either, a sidebar pops up showing the exact post where the information comes from.

An example results page for Reddit Answers.

This is from a test version of Reddit Answers, so things might look a tad different when you try it. But you can get the gist.

A Reddit site search for the same query on Google, on the other hand, just gave me a standard long list of links to evaluate and click through. On its face, Google’s version, which doesn’t have the summaries or bullet points, isn’t as immediately informative as Reddit Answers. But personally, I’m pretty skeptical of AI summaries anyway — they can make some pretty bad errors! — so for me, the Google results are still perfectly fine as a way to explore some potentially useful posts on Reddit. 

A screenshot of a Google search for “site:reddit.com + tips for flying with a baby for the first time”

My Reddit site search on Google.

Reddit Answers can also pick up on things happening on Reddit within minutes, according to spokesperson Tim Rathschmidt, so there’s also a chance responses to your queries will have timely results. I haven’t had much luck with that; on Friday, for example, I asked “who won yesterday’s NFL game,” and instead of pulling up answers about Thursday’s Packers-Lions game, it included details on the Eagles-Ravens game that happened on December 1st.

That might have just been a quirk of the product not being fully ready yet. In my testing, though, Reddit Answers has been much more useful as a jumping off point to dig into a more general topic on Reddit rather than an up-to-the-minute s

you are viewing a single comment's thread
view the rest of the comments
[–] sugar_in_your_tea@sh.itjust.works 2 points 2 weeks ago (1 children)

Would you have said the same about early computers, when they were less effective at solving problems than a human?

The problem isn't AI, but the crazy investment into hype, which results in companies forcing it where it doesn't belong. Total energy use of data centers (not just AI, all data centers, meaning pretty much every website) is about 1% of global electricity, which is about half the energy use of home electronics. I don't know what share of that is AI, but I would be surprised if it's even half of the data center usage.

Yeah, it's wasteful now, but just like early computers, it has the potential of being very effective. Some areas I'm interested in are medical diagnosis and improving accessibility of technical documentation. Doctors and nurses already use algorithms for diagnosis ie. flowcharts), and studies indicate that they make medical professionals more effective, and AI can add to that. I don't think it's nearly as useful for the average person especially due to the hallucination problem, but targeting non-experts can help refine things for experts.

I also generally avoid AI because I'm concerned about not catching hallucinations on things I'm unfamiliar with, and wasting time on correcting things on things I am familiar with. However, I see promise in it, so I'm interested in following developments in the field.

[–] nokturne213@sopuli.xyz 0 points 2 weeks ago (1 children)

Would you have said the same about early computers

If we were in a worldwide drought with temperatures increasing to record levels with increasingly violent storms, yes. We need to scale back on power and water consumption, not increase it.

And straining at gnats isn't going to get us there. AI uses <1% of global energy usage, it's not what's causing the problem. Much bigger contributors to CO2 pollution are personal vehicles, commercial vehicles, and industry (esp. producing plastic crap), not data centers.