builder
Joined January 2025
- Tweets 622
- Following 1,157
- Followers 115
- Likes 506
27 Photos and videos
Pinned Tweet
Check out some great AI Spaces at the
Build-small Hackathon
hosted by @huggingface and @Gradio
Here's mine:
huggingface.co/spaces/build-…
1
1
1
496
Couldn't have wrote a better blog in favor of Open Source AI or Hermes Agent's self improvement loop!
Thanks for the strong words in support of taking back the power to the individuals and businesses who really ought to own their stack.
43
57
972
72,900
broadfield-dev retweeted
RIP Oliver Tree 💔🕊️
246
1,228
28,419
1,468,958
Can my last minute SEO land me in the AI description of the Search Engines within 25 hours?
@gradio Build Small Hackathon
build-small-hackathon-overth…
1
178
Check out 600 AI submissions to the
Build Small Hackathon
huggingface.co/build-small-h…
13
Overthink it
Check out some great AI Spaces at the
Build-small Hackathon
hosted by @huggingface and @Gradio
Here's mine:
huggingface.co/spaces/build-…
18
Check out some great AI Spaces at the
Build-small Hackathon
hosted by @huggingface and @Gradio
Here's mine:
huggingface.co/spaces/build-…
1
1
1
496
my agent loop has always just run in a loop until it was done with it's task. I don't understand toggling on and off /goal?
Jun 14
Grok Build now has 57 slash commands. Many of them are really useful and unique. The only command i'm still looking forwar to is /goal.
12
What does it do when it's not achieving a /goal @xai?
Jun 14
Grok Build now has 57 slash commands. Many of them are really useful and unique. The only command i'm still looking forwar to is /goal.
10
broadfield-dev retweeted
Two PII models this weekend, opposite ends of the spectrum.
A 3.3B generative model I RL-trained (GRPO) until it emits 100% valid JSON across 20 languages, 69 entity types.
And a 33M one, 37MB, that runs a realtime privacy filter on your iPhone = MLX
Both open, Apache 2.0. 👇
14
22
223
13,277
broadfield-dev retweeted
Jun 14
There’s not that much Latin online. AI still struggles to understand it. This gives us a wonderful opportunity to poison the internet with terabytes of fake Latin, which AIs will train on. Then we humans can use real Latin to communicate amongst ourselves in secret.
146
163
3,835
45,174
Idea:
Only train the model how to code.
Don't train it to do anything else.
Run it on a cellphone.
1
25
1
54