I agree with your post, but in practice, couldn't you get back that efficiency by setting T = UInt8Array? That is, write your stream to send / receive arrays.
My reference point is from a noob experience with Golang - where I was losing a bunch of efficiency to channel overhead from sending millions of small items. Sending batches of ~1000 instead cut that down to a negligible amount. It is a little less ergonomic to work with (adding a nesting level to your loop).
Sex work shouldn't be shunned, but it's not a normal profession either. Mental health, addiction and abuse is just as much of a problem online and in countries where prostitution is legal and normalized.
What the surgery actually does is fix 2 disks of your spinal column against each other. It lowers the pain from torture to tolerable and reduces various risks. Also: you won't be so much as sitting up for months. I don't think many people will call that repair. Perhaps mitigation.
I think you are both talking about slightly different things:
* Herniated disk in the spine
* A "hernia": is the abnormal exit of tissue or an organ, such as the bowel, through the wall of the cavity in which it normally resides.[
If you think of your abdomen as a bag full of tubes, a fistula is a hole in the tube that connects to something else. A hernia is a hole in the bag, that the tube can poke through.
Uber's in a business where you have some amount of network effect - you need both drivers available using your app, as well as customers hailing rides. Without a sufficient quantity of either, you can't really turn a profit.
LLM providers don't, really. As far as I can tell, their moat is the ability to train a model, and possessing the hardware to run it. Also, open-weight models provide a floor for model training. I think their big bet is that gathering user-data from interactions with the LLM will be so valuable that it results in substantially-better models, but I'm not sure that's the case.
in contrast, a toddler equipped with an ion thruster & a modest quantity of xeon propellant could achieve enough delta-v to attain cheetah-escape velocity, provided the initial trajectory during the first 31 hours of the mission was through a low-cheetah-density environment
That initial trajectory also needs to go through a low air density environment. At normal air density near the surface of the Earth that ion thruster could only get a toddler up to ~10 km/h before the drag force from the air equals the thrust from the ion thruster.
The only way that ion thruster might save the toddler is if it was used to blast the cheetah in the face. It would take a pretty long time to actually cause enough damage to force the cheetah to stop, but it might be annoying enough and/or unusual enough to get it to decide to leave.
> low air density environment. At normal air density near the surface of the Earth that ion thruster could only get a toddler up to ~10 km/h
agreed. this also provides an explanation for the otherwise surprising fact that prey animals in the savannah have never been observed to naturally evolve ion thrusters.
Looks pretty cool. After checking out spec.md, one thing I might suggest is using UUIDs for external codec ids. There's a lot of codecs/formats out there, and limiting to a u8 might lead to collisions.
It might also be nice to provide a mechanism to advertise the required codecs toward the beginning of the stream, in case the consumer does not have the necessary codecs and wishes to abort the transfer.
Good catch — you’re absolutely right to call that out.
The current u8 codec ID is mainly there to keep the block header very small and fast to parse, but it’s not meant to be the global identifier. The idea is to map that ID to something globally unique (most likely a UUID) through the plugin/manifest layer, so we can avoid collisions without bloating the on-disk format.
I also like the suggestion about advertising required codecs early in the stream. That would make it much nicer for a reader to fail fast if it doesn’t support something, especially for streaming use cases. We’re exploring adding a small capability section near the beginning for exactly that reason.
Since the format is still experimental, this kind of feedback is really helpful before we lock things down.
Thank you for your message. English is not my native language, so I sometimes use translation tools. I will try my best to reply to you in more direct and understandable language. Thank you for your patience.
If you want to use smaller ids within the stream, that capability section would seem to be the natural place to map from global codec-uuid to file-local u8 identifier
The youtube video currently has ~1.4 million views. Colbert averages 2 to 3 million television viewers per night, plus some number of youtube views that I haven't looked up the stats for.
That is, this interview has been seen by fewer people than it would have, had it been on television.
3.1 million views as the sun goes down here in Tennessee. I'll concede that's not significantly "surpass" but it's still approximately equal, and may have been more people than saw the actual broadcast.
reply