#zippy — Public Fediverse posts
Live and recent posts from across the Fediverse tagged #zippy, aggregated by home.social.
-
My Life with Puppets: The Ronnie Le Drew Story
RIP Ronnie Le Drew the voice of Zippy.
-
Not sure what that is that showed up in this photo- didn't see anything- was just star gazing and contemplating- #zippy #star #lights #night #sky #observations #photography #clouds
-
Remember the baby rabbit that wouldn't leave the garden? It's round the back of the house, by the kitchen window.
#Zippy -
Caught the express at Times Square today and boy, was he down on getting us there right quick. Zippy trip this morning, the skipped stations going by in blurs.
#blurry #commute #today #NYC #subway #NewYorkCity #work #WorldTradeCenter #station #7WTC #express #take the A #train #zippy #trippy #ride #fast #trip #TimesSquare #conductor #quick #driver #speed #morning #coffee no #decaf -
Starting February, Gmail will require you to authenticate outgoing email, avoid sending unwanted or unsolicited email (meaning to use SPF, DKIM, DMARC and some more) if you're a "bulk sender" sending over 5k emails/day to Gmail.
Few points I've learned when checking the requirement:
- only messages to personal Gmail accounts count towards the 5k/day limit- emails from both @example.com and @foo.example.com count towards the "example.com" limit
- once a bulk sender, always a bulk sender
- you can use DMARC with p=none policy which basically means "I have a DMARC record but...", it's a good starting point and makes sense when used with rua ("send me aggregated reports") to check how you're doing before switching to p=quarantine ("your" messages that don't pass SPF/DKIM checks end up in spam folders") or p=reject (such messages are rejected)
- enforcement for bulk senders will be gradual and progressive: small percentage of non-compliant emails will see temporary errors in February, in April In April 2024, Gmail we’ll start rejecting a percentage of non-compliant email traffic, gradually increasing the percentage of rejected emails
Another requirement is a one-click unsubscribe which bulk senders have until June 1, 2024 to implement.
There's a lot of info out there and I found these two pages the most helpful: this one is a bit higher-level https://support.google.com/a/answer/81126?#zippy=%2Crequirements-for-all-senders
and this one has all the details:
https://support.google.com/a/answer/14229414 -
@hack_lu @Jacob the recent detector from #contentatscale beats #zippy in accuracy but alas not in speed:
-
For you #firefox users, now you can join in the fun by automatically fading out suspected LLM-generated content as you browse. This version is running the new #zlib version of #ZipPy, so it's very faster, but 2% less accurate. Feedback is welcome 🤗
https://addons.mozilla.org/addon/ai-noise-cancelling-headphones/
#browser #extension #ai #detector #LLM -
We have a new winner in terms of #AI content detection accuracy! The https://contentatscale.ai/ai-content-detector/ #detector is impressive in its accuracy, despite being more than 200x slower than #ZipPy. Also shown in this new #ROC chart is the comparison between #zlib at maximum compression setting vs. #LZMA at a low preset (2). zlib is faster (especially in the #browser with native #CompressionStream support), but not quite as performant.
-
Based on the open-source #ZipPy #AI detector: https://chrome.google.com/webstore/detail/ai-noise-cancelling-headp/okghlbkbacncfnfcielbncabioedklcn
Feel free to check the source if you're concerned, but 0 tracking, all local, no networking calls out: https://github.com/thinkst/zippy/tree/main/inch -
#Ziggy by #TomWilson for 1985-2-15.
-
Check out the new #LLM #AI text detection tool I've been working on at @ThinkstCanary : #ZipPy!
https://blog.thinkst.com/2023/06/meet-zippy-a-fast-ai-llm-text-detector.html -
