Follow us to see the best and most innovative features on the internet. @FeatureTweeter with a #featurerequest for a brand you love!
Ann Arbor, MI
Joined October 2021
- Tweets356
- Following568
- Followers73
- Likes580
Feature Tweeter retweeted
Let's all hope that the Siri implementation is like @openclaw
Feature Tweeter retweeted
Hey @GeminiApp — one thing’s missing: a copy button in the Gemini app.
I’ve tried everything — long-press, menus, export, nothing feels clean or built-in.
Can you please add a native copy feature so it’s easier for users to share and reuse output?
#Gemini #FeatureRequest #AItools
@GoogleDeepMind
@Google
Feature Tweeter retweeted
App:@Twitter
Feature Request: Allow users to set a different avi & background image for dark mode followers & light mode followers
#featurerequest #Twitter #aesthetic
Feature Tweeter retweeted
This video paired to this one of Tolkien reading the Ride of the Rohirrim is perfection
Feature Tweeter retweeted
If I had a huge Twitter account, I would make physical awards and send them to my best followers.
“Reply Guy of the Year”
“Best IRL Connection”
“Best New Follower”
“Best Ratio on Matt’s Tweets”
“Best Troll”
Etc. – at Sioux Falls, SD
Feature Tweeter retweeted
iMessage on Android?
Thanks to @nothing it's about to become a reality
The company will soon launch "Nothing Chats" which will bring the blue message bubble to Android users
theverge.com/2023/11/14/2396…
Feature Tweeter retweeted
🆕 Coming soon to @googledrive: A filter view that lets you see which files or folders have been shared with specific people or groups. Learn more → goo.gle/40MrEVd
Feature Tweeter retweeted
AI is now inside Google Maps!
Prepare to be blown away by these 5 mind-blowing AI features!
(number 2 is my favorite)
Feature Tweeter retweeted
Hey Mindy! That does sound like an interesting idea. You should suggest it by visiting crumblcookies.com/contact and "Make a suggestion"!
Feature Tweeter retweeted
An interesting thing I have observed w/ DАLL-Ε 3 is that it emulates compression artifacts, similar to a strongly compressed jpg or video file.
DАLL-Ε 3 itself outputs lossless png files, so it wouldn’t come from that.
Yet because of its training data it “learned” that darker regions are often noisy / blocky.
It only does it for “photorealistic” images, not artwork etc.
The lighting reminds me a lot of the way modern movies or TV shows are lit, which leads me to believe it was likely trained on movie / TV show footage, which tend to be noisy in darker scenes.
Feature Tweeter retweeted
Probably the first entity to feel the fire of LLM eating the world 🔥
This was analyzed very well by a recent paper
Paper Link - arxiv.org/abs/2307.07367
🔥⚠ Large Language Models impacting StackOverflow🔥⚠
📌 The number of weekly posts decreases at a rate of about 7,000 posts each year from 2016 to 2022. In the six months after the release of ChatGPT, the weekly posting rate
decreases by around 20,000 posts.
📌 Compared to the pre-ChatGPT trend, this decrease represents more than five years worth of deceleration in just half a year.
📌 Very insightful paper - "Are Large Language Models a Threat to Digital Public Goods? Evidence from Activity on Stack Overflow"
Feature Tweeter retweeted
StackOverflow is laying off 28% of its workforce.
This may be the first large layoff directly due to AI:
> people asking ChatGPT instead of StackOverflow
> usage & ad revenue declines
> having to lay people off to stay profitable / survive
Readers added context they thought people might want to know
This chart shows a significant drop in traffic before ChatGPT's release, November 30, 2022.
StackOverflow has confirmed that inaccurate data regarding their website traffic is circulating online. This year, they have measured an average of ~5% less traffic compared to 2022: stackoverflow.blog/2023/08/08/ins…