okaris

Unfortunately the bar now is very very low. Thanks for the response!

thanks for doing some real journalism here, but why aren’t the sources named? without that, it’s still something we have to take on trust, not much stronger than the original rumor.

it was wild to watch this come alive. workflows are cool, but here you just do stuff. the bot can use every app on inference.sh and even make new ones on the fly. video trimmed to skip wait times. the best part? it all runs local.

workflows? where we're going, we don't need roads.

I've found that the "ceefax" keyword looks great on flux ultra.

(CEEFAX was an early text service that broadcast on BBC channels in the empty parts of a TV signal)

Happy new year friends. I have big plans and high hopes for 2025

@dubmood.bsky.social Huge fan since the early 2000s when I first discovered ChipDisk #1 and #2. A few weeks ago, I came across the Cosmoteer soundtracks, and it brought me back 20 years. If it’s alright, could I DM you about commission work?

iykyk

Came across this very creative ai project today: hexagen.world where thousands of people generate world tiles together.

For my next project, I’m adding a built-in self-destruct function. Sunsetting a project isn’t as small a task as it seems, and trying to support a few remaining users in old projects has become a distraction.

Might also prevent AI from taking over

Been diving into Go recently. Channels and goroutines felt a bit weird at first, but once they clicked, everything came together so smoothly. Loving how simple and intuitive it all feels.

It needs a do over but can’t DM. You can send over a translation file if you want

I’ll take care of the Turkish

Now it’s just up to us to create good content :)

FYI, here's the entire code to create a dataset of every single bsky message in real time:

```
from atproto import *
def f(m): print(m.header, parse_subscribe_repos_message())
FirehoseSubscribeReposClient().start(f)
```

Hey Jeremy! Great to see you here

👋🏻

I believe there’s often a lot of room for optimization before jumping into creating a CUDA kernel. Many models and libraries are overly bloated in an attempt to cater to broad use cases.

Our lives are shaped by a series of improbable events. Which means the next miracle is just as likely.

Thank you! 🍻

Trying to figure our how I missed my handle again now on bluesky 😂 Someone keeps getting @okaris and does nothing with it.

@okaris.bsky.social hi

inference

Hi BlueSky