Kagi search engine

339 readers

4 users here now

A community to discuss the innovative paid Kagi search engine and related topics.

Kagi Inc. is a company created with the mission to humanize the web. Our goal is to amplify the web of human knowledge, creativity, and self-expression.

https://kagi.com/

Rules: Be moral.

Note: This community is not affiliated with Kagi Inc.

founded 2 years ago

MODERATORS

Atemu@lemmy.ml

Kagi Blog: Introducing SlopStop: Community-driven AI slop detection in Kagi Search (blog.kagi.com)

submitted 4 months ago by lazycouchpotato@lemmy.world to c/kagi@lemmy.ml

8 comments fedilink hide all child comments

you are viewing a single comment's thread
view the rest of the comments

[–] Novocirab@feddit.org 3 points 4 months ago (1 children)

On the one hand, that's nice for improved Kagi results. On the other hand, I dislike how community efforts will end up being owned by a single company.

[–] lazycouchpotato@lemmy.world 6 points 4 months ago (2 children)

They addressed it under:

Access to the database will be shared soon, you can express interest here if you’d like to receive updates.

I'm not sure what the criteria will be for receiving access to the database.

[–] Novocirab@feddit.org 2 points 4 months ago* (last edited 4 months ago) (2 children)

Interesting, thank you. From the text here,

"As we collectively identify and validate slop across the web, Kagi's SlopStop initiative will help transform those insights into a comprehensive, structured dataset of AI slop – an invaluable resource for training AI models.

Access to the database will be shared soon. Use this form to express your interest if you’d like to receive updates.

especially the part "an invaluable resource for training AI models", and the absence of any community-focused language, and the fact that Kagi iirc is still operating at losses and are looking for ways to become profitable, I fear it will be essentially commercial. But who knows, and even if it will be so, that still is not to say it's all bad.

[–] thibaultmol@en.osm.town 2 points 4 months ago

@Novocirab
Small side note:
Kagi isn't at a loss, they're profitable. They said so earlier this year
@lazycouchpotato

[–] AmbitiousProcess@piefed.social 1 points 4 months ago

and the fact that Kagi iirc is still operating at losses and are looking for ways to become profitable

Kagi is already profitable. A good sign, imo.

See: https://blog.kagi.com/what-is-next-for-kagi where they say "We are also thrilled to report that we have achieved profitability" on May 30 of last year.

As for the AI models, I see this as only really being a good thing. Either existing AI models (i.e. LLMs) get trained to produce less sloppy content, and thus less content generated by them is actually inaccurate and damaging to the web's integrity as a source in the first place, or it's used to train AI models made to detect slop, like they've said they want to do, in which case it becomes easier to block out low-quality slop without even needing human reports for the majority of site blocks.

[–] cecilkorik@piefed.ca 1 points 4 months ago

Nobody is sure yet, but it doesn't matter because access is not even close to the same thing as ownership. We need to stop letting people own our crowdsourced data. They are not "adding value", they are parasitically feeding on us.