XL was basically an experiment on a 2.1 architecture with some tweaks but at a larger image size... hence the XL but it wasn't really an evolution of the underlying architecture which is why it wasn't 3.0 or even 2.5 it was "bigger" lol
The speed at which the models are progressing and new optimizations for reducing hardware requirements makes $20/mo per user not too bad. They'll be able to reduce those costs pretty soon in my opinion, and the users will continue to use the service for years to come.
The flip side to that is it's going to be easier to self-host a coding LLM. That's free for the user (if they already have a good enough GPU) and it gives much better flexibility.
That's a good point, the large corporation model vs smaller specialized finetuned models are vying for dominance right now, I'm interested to see which paradigm we end up with.
Won't be winner takes all. The people who use local models will have requirements or desires that lean them that way. The rest will use the easier to stand up large corp offerings.
Only way this changes is if there is a very large performance gap between the two. Small ones just won't matter.
In addition to sitting vs non-sitting periods, I want the study to address the types of activities that people are doing during the sedentary periods. For instance, would sitting and watching TV be as detrimental as sitting and working at a desk?
I guess we gotta wait a few more decades for those results, and maybe they can incorporate a combination of accelerometer data along with screen time.
^ This was important for me. I think it may have something to do with the separation of environments from bedroom-office (primed for resting) and an actual workspace (productive space).
Beyond the high price to just purchase a car, Singapore freeways are loaded with tolls which make the price to operate vehicles even higher.
The flip side, the roads/infrastructure/landscaping of Singapore are the best I've ever seen. Modern and pristine, at least we know those funds are being put to good use.
Oh? I've had trouble with Overcast. It often exits when I launch it, usually this happens after I haven't used the app for a few days. The app works when I launch it a second time. When it works, it takes a long time, e.g., a full minute before the app is ready to accept input and start a podcast. I haven't taken the time to look at it closely or request support. There's plenty of storage available, so it's not a "disc full" error in disguise. I've gone back to the Apple iOS app.
My use case is a few regular podcasts, and many (~100) podcasts where I listen to only one episode. I like to keep the infrequently accessed podcasts around so I can remember what I listened to, and in case there's another episode I want.
Also it doesn't sync with iCloud, but maybe that's been added?
