@cdarwin@c.imUnlike keyword search,
semantic search lets you search using natural language.
It looks beyond exact matches to understand the meaning and intent behind your query.
This means it can surface relevant precedents even when they're phrased differently
—something keyword searches often miss.
Semantic search is currently available through an API, but we're already working to bring it to the website—stay tuned!
And don't worry, keyword search isn…
@ErikJonker@mastodon.socialSometimes i have the feeling that AI results in Google Search are just a placeholder. Results are often awful and incomparable with using the paid version of Gemini 3 with the same query. Google is betting that models will improve and become cheaper so that they can offer a better experience in the future.
#AI #Search
@outer@mas.toCustom self-leveling, poured polyurethane art floors. Cool video.
I found others, e.g., on youtube:
"epoxy flooring 3d"
https://www.youtube.com/results?search_query=epoxy flooring 3d
@CerstinMahlow@mastodon.acm.orgYesterday, I observed a friend trying to show me something he saw on facebook recently. Every single search he tried resulted in first switching from facebook to “Meta”—whatever that is—without him noticing for quite some interactions, the search field was still there. Sending a query always opened the Meta AI answering the assumed question (but he was actually sending a query as he was looking for an image). He got angry, I got amused. He found the image eventually, but what a wacky UI!
@grahamperrin@bsd.cafe@… ha, thanks … the bears page was amongst the top search results at <https://www.startpage.com/do/dsearch?query="d…
@bourgwick@heads.socialsome excellent notes on the recent stash of boredoms recordings surfaced by @…. https://archive.org/search?query=subject:"boredoms%2…
@cjust@infosec.exchange
@midtsveen@social.linux.pizzaThis rock wrote some seriously based shit you should definitely read!
🔗 #Anarchism
@bobmueller@mastodon.world"Hey, let's get some Chinese characters tattooed on our body without knowing any Chinese language!"
https://www.youtube.com/@ChinesewithJessie/search?query=Tattoo
@cdarwin@c.im
@michabbb@social.vivaldi.net📈 #LogsQL query language provides fast full-text search, advanced analytics, and data extraction/transformation at query time. Can be combined with Unix tools like grep, less, sort, and jq for log analysis.
🎯 Optimized for high cardinality fields like trace_id, user_id, and ip addresses. Supports logs with hundreds of fields (wide events), multitenancy, out-of-order ingestion, live taili…
@tiotasram@kolektiva.socialImagine ChatGPT but instead of predicting text it just linked you to the to 3 documents most-influential on the probabilities that would have been used to predict that text.
Could even generate some info about which parts of each would have been combined how.
There would still be issues with how training data is sourced and filtered, but these could be solved by crawling normally respecting robots.txt and by paying filterers a fair wage with a more relaxed work schedule and mental health support.
The energy issues are mainly about wild future investment and wasteful query spam, not optimized present-day per-query usage.
Is this "just search?"
Yes, but it would have some advantages for a lot of use cases, mainly in synthesizing results across multiple documents and in leveraging a language model more fully to find relevant stuff.
When we talk about the harms of current corporate LLMs, the opportunity cost of NOT building things like this is part of that.
The equivalent for art would have been so amazing too! "Here are some artists that can do what you want, with examples pulled from their portfolios."
It would be a really cool coding assistant that I'd actually encourage my students to use (with some guidelines).
#AI #GenAI #LLMs
@nelson@tech.lgbtRecently reminded of Straight to Hell, the 1975—20?? zine featuring porny gay stories that were often real life narratives. Transgressive and important documentation. The Internet Archive has PDF scans of 5 edited book collections!
Harvard man, Harvard man,
blond for no reason. I'd
sure as shit quit the Quad
for him, scratch every dance
from my card for him; kiss
sunshine goodbye to open
his fly; trade clothes
for sheets, miss other meets
for him, for him
@tomkalei@machteburch.socialOK, but the results are still just paper links, not what the LLM thinks about them. So the only thing that changed is that it is much much slower now?
Maybe the LLM just formulates a query to the old scholar? Too many layers of "oh how can we build an LLM into this"...
OK, nobody in #math uses Google Scholar like this anyway. We only use it look at publication records of people and do reverse citation search (which papers cite this one I have).
