Well shucks, I guess Samsung lied to me! I see adaptive brightness was released as part of Android Pie in 2018.
Yeah, I was thinking it could be machine learning in that it takes the average of all your changes over time and the different ambient light levels.
But deffo no need for neural networks.
I totally agree with you.
However there is one smart feature Samsung has that I like. The screen brightness auto adjusts based on the ambient light, but if I change that automatic brightness (I prefer the screen darker) it will remember that and consistently adjust the brightness.
I’m not sure it’s really an AI feature…
Take the ambient light level (lux).
Set brightness to 5.
Log that the user has made it 1 level or 10% darker.
Next time it senses the same lux level, set the brightness 1 level lower
The UI on the app is really disingenuous, it looks like there is no option to change until you click on the individual descriptions.
It’s US English I think …
The other one that I notice is “write your representative”, instead of “write to your representative”
I didn’t even know this was a thing!
Lol it’s clearly from the statue level in Goldeneye…
Exercise bike or rower perhaps? But I’m a fan of cardiovascular exercise.
You could perhaps get some nice running shoes and weights.
I’m pretty sure the USA refers to the model*. The manufacturer may have a Chinese or International version of each phone.
I totally agree with you, I’d much rather read some bullet points then sit through a video.
I think YouTube automatically generates transcripts already, and GPT can summarise text easily.
We’ll get an AI to caption and then summarise the transcript of the video that’s 8 mins long because the user (understandably) wanted to monetize the video.
I wonder if this was Google’s plan all along …
This is neat solution, you’ll lose the laptop WiFi internet access though.
You can set the pi to connect to the phone WiFi by editing files on the SD card (using your laptop)
https://desertbot.io/blog/headless-raspberry-pi-4-ssh-wifi-setup-64-bit-mac-windows
Pi4 with 2TB SSD running:
HDMI cable straight to the living room Smart TV (which is not connected to the internet).
Other devices access media (TV shows, movies, books, comics, audiobooks) using VLC DLNA. Except for e-readers which just use the Calibre web UI.
Main router is flashed with OpenWrt and running DNS adblocker. Ethernet running to 2nd router upstairs and to main PC. Small WiFi repeater with ethernet in the basement. It’s not a huge house, but it does have old thick walls which are terrible for WiFi propogation.
I’m still on an early gen Chromecast and never see any adverts. Very occasionally crashes (once a fortnight).
If I understand correctly Pi hole works as a DNS service so I don’t think the bandwidth (100mb) is an issue.
You’re not routing all the network data through the Pi.
The issue is wether the Pi can run fast enough to process all your lookup requests.
Gallery link only shows one picture for me in Sync for Lemmy
Slightly off topic, but one major benefit for me with Chromecast and keeping the TV offline is software bloat.
Our old Samsung TV running tizan had so many shitty updates that slowed the UI down to what felt like over a second delay when using the remote.
I don’t know if you’ve already looked but this site is my favourite for this kind of question.
Has the ability to filter by licence and platform too. https://alternativeto.net/software/ourgroceries/
So I was trying to think how I’d implement it, and I agree if it’s simple then it really only needs to set the brightness level once, then remember if the user adjusts it, and reuse that adjustment for every lux reading.
Hence the example I gave: