𝚐π”ͺ𝟾𝚑𝚑𝟾 pfp

𝚐π”ͺ𝟾𝚑𝚑𝟾

@gm8xx8

138 Following
69488 Followers


𝚐π”ͺ𝟾𝚑𝚑𝟾 pfp
𝚐π”ͺ𝟾𝚑𝚑𝟾
@gm8xx8
LONG LIVE OPEN SOURCE
0 reply
6 recasts
22 reactions

𝚐π”ͺ𝟾𝚑𝚑𝟾 pfp
𝚐π”ͺ𝟾𝚑𝚑𝟾
@gm8xx8
Still here, just not casting as much. I’m tracking AI on Farcaster daily and inviting signal-heavy casters who show consistent interest. Modbot was a temporary solution and it’s down, so I’m working on a better way to keep the channel focused. For now, it’ll stay invite-only. The channel’s been shaped with intention, and I believe a curated, high-signal AI space is something farcaster will need as it matures. That said, the beauty of farcaster is that there are plenty of channels…everyone can find their lane.
0 reply
0 recast
0 reaction

𝚐π”ͺ𝟾𝚑𝚑𝟾 pfp
𝚐π”ͺ𝟾𝚑𝚑𝟾
@gm8xx8
Soft no then, lol. Had this bookmarked, looks like I missed the context. Still, you’re always welcome on my side of Farcaster.
0 reply
0 recast
1 reaction

𝚐π”ͺ𝟾𝚑𝚑𝟾 pfp
𝚐π”ͺ𝟾𝚑𝚑𝟾
@gm8xx8
Did it bring any new users to Farcaster? Just curious if it’s helped get more attention here. Doesn’t look that way imo but i could be wrong.
1 reply
0 recast
1 reaction

𝚐π”ͺ𝟾𝚑𝚑𝟾 pfp
𝚐π”ͺ𝟾𝚑𝚑𝟾
@gm8xx8
we are not the same
7 replies
64 recasts
317 reactions

𝚐π”ͺ𝟾𝚑𝚑𝟾 pfp
𝚐π”ͺ𝟾𝚑𝚑𝟾
@gm8xx8
appreciate it. glad to have you in thereβ€”looking forward to your takes.
0 reply
0 recast
1 reaction

𝚐π”ͺ𝟾𝚑𝚑𝟾 pfp
𝚐π”ͺ𝟾𝚑𝚑𝟾
@gm8xx8
I blocked alex not bc he posted in this channel but bc he pulled this.
1 reply
0 recast
1 reaction

𝚐π”ͺ𝟾𝚑𝚑𝟾 pfp
𝚐π”ͺ𝟾𝚑𝚑𝟾
@gm8xx8
appreciate that v. might want to let dwr know i’m also a great follow on x with under 10k.
1 reply
0 recast
5 reactions

𝚐π”ͺ𝟾𝚑𝚑𝟾 pfp
𝚐π”ͺ𝟾𝚑𝚑𝟾
@gm8xx8
You ask me who to follow in AI, but don’t even follow meβ€”after I’ve spent two years sharing insights and building a presence on Farcaster. Then, there’s another person who literally copied my following list, recommended others follow me, but when I tried to return the favor? Realized I’m blocked. I guess they just wanted to give the impression they are in the know. I see why so many AI folks have churned ☹︎ I’ll still check in and post occasionallyβ€”this isn’t me quitting Farcaster or anything like that, just sharing some πŸ’­ Attached below is one example, but I’ve received plenty of similar messagesβ€”some harsher towards farcaster, all eye-opening, all from AI folks. I also made sure they knew I’m still here, just not posting as often. My experience hasn’t been all badβ€”many from Farcaster have reached out to say they only know what they do about AI because of me and have shared their gratitude. That support is why I still return.
11 replies
1 recast
17 reactions

𝚐π”ͺ𝟾𝚑𝚑𝟾 pfp
𝚐π”ͺ𝟾𝚑𝚑𝟾
@gm8xx8
https://warpcast.com/gm8xx8/0xb8d6774c
1 reply
0 recast
7 reactions

kevin j pfp
kevin j
@entropybender
easily @gm8xx8
1 reply
1 recast
4 reactions

𝚐π”ͺ𝟾𝚑𝚑𝟾 pfp
𝚐π”ͺ𝟾𝚑𝚑𝟾
@gm8xx8
sharing for those unfamiliar w/ lerobot https://warpcast.com/gm8xx8/0x5487dd2b
1 reply
0 recast
1 reaction

𝚐π”ͺ𝟾𝚑𝚑𝟾 pfp
𝚐π”ͺ𝟾𝚑𝚑𝟾
@gm8xx8
πŸ‘
1 reply
0 recast
1 reaction

𝚐π”ͺ𝟾𝚑𝚑𝟾 pfp
𝚐π”ͺ𝟾𝚑𝚑𝟾
@gm8xx8
Physical-Intelligence Ο€β‚€ is now open source. > Ο€β‚€ and Ο€β‚€-FAST Model Weights > source code for model architecture, on-robot Inference, and fine-tuning > pre-trained fine-tuned checkpoints robotics got another boost βœ”οΈ πŸ”—: https://github.com/Physical-Intelligence/openpi https://warpcast.com/gm8xx8/0xce446768
2 replies
1 recast
16 reactions

𝚐π”ͺ𝟾𝚑𝚑𝟾 pfp
𝚐π”ͺ𝟾𝚑𝚑𝟾
@gm8xx8
FAST is a robot action tokenizer that simplifies and speeds up robot training. It enables: > 5x faster training compared to diffusion models. > Compatibility with all tested robot datasets. > Zero-shot performance in new environments, including the DROID dataset, successfully controlling robots in various settings with ease. > Simple autoregressive VLAs that match diffusion VLA performance. > Mixed-data VLA training, allowing integration of non-robot data like web data, subgoals, and video prediction. FAST compresses actions using discrete cosine transform, reducing redundancy and enabling efficient VLA training on high-frequency tasks. It scales to complex robot tasks with simple next-token prediction, converging in days instead of weeks. A pre-trained FAST tokenizer based on 1M robot action sequences is available on Hugging Face, working across various robots and supporting mixed-data VLA training. https://huggingface.co/physical-intelligence/fast
0 reply
4 recasts
19 reactions

𝚐π”ͺ𝟾𝚑𝚑𝟾 pfp
𝚐π”ͺ𝟾𝚑𝚑𝟾
@gm8xx8
cc @sdv.eth
0 reply
0 recast
0 reaction

𝚐π”ͺ𝟾𝚑𝚑𝟾 pfp
𝚐π”ͺ𝟾𝚑𝚑𝟾
@gm8xx8
> literally all the other content is 3 day old twitter from KOLs exactly my point. https://warpcast.com/gm8xx8/0x81c0d34a
0 reply
0 recast
4 reactions

𝚐π”ͺ𝟾𝚑𝚑𝟾 pfp
𝚐π”ͺ𝟾𝚑𝚑𝟾
@gm8xx8
(q‒̀ᴗ-)βœ§βŒ’β™‘* anytime
0 reply
0 recast
1 reaction

𝚐π”ͺ𝟾𝚑𝚑𝟾 pfp
𝚐π”ͺ𝟾𝚑𝚑𝟾
@gm8xx8
DeepSeek releases Janus-Pro-7B and Janus-Pro-1B πŸ‹
1 reply
3 recasts
21 reactions

𝚐π”ͺ𝟾𝚑𝚑𝟾 pfp
𝚐π”ͺ𝟾𝚑𝚑𝟾
@gm8xx8
|οΏ£οΏ£οΏ£οΏ£οΏ£οΏ£οΏ£οΏ£οΏ£οΏ£οΏ£οΏ£οΏ£l | everytime I open this app it l | feels like a step back in l l time. l |οΌΏοΌΏοΌΏοΌΏοΌΏοΌΏοΌΏοΌΏοΌΏοΌΏοΌΏοΌΏοΌΏ| \ (β€’β—‘β€’) / \ / β€”β€” | | |_ |_
4 replies
12 recasts
28 reactions