@bambambebbam everything in crypto is about attention, and the attention is gone if you are about token
if you about project itself than the tech is outdated and the approach as well
been a while
spent some time working irl and honestly didn’t expect this account to still have people around.
might disappoint someone, but not sure sam itself makes sense to keep pushing as a “project” anymore, but I do miss building in ai, and this place still feels like a solid starting point
I still have all the old sam holder snapshots too, so if I build something new, old holders probably won’t be forgotten
doesn’t mean new token tomorrow or any of that. just finally got free time again and the urge to build came back
been a while
spent some time working irl and honestly didn’t expect this account to still have people around.
might disappoint someone, but not sure sam itself makes sense to keep pushing as a “project” anymore, but I do miss building in ai, and this place still feels like a solid starting point
I still have all the old sam holder snapshots too, so if I build something new, old holders probably won’t be forgotten
doesn’t mean new token tomorrow or any of that. just finally got free time again and the urge to build came back
Good morning. Fresh commit is in, all tests passing on GitHub.
Token rebrand is live on @dexscreener with the new GitHub link and a dedicated website just for OpenSAM.
One-liner install is here: rent a vps or use your own device, run one command, do a quick interactive setup, and you’ve got a 24/7 agent with persistent memory.
OpenSAM = $SAM
https://t.co/7xPsSTbAXS
Good morning. Fresh commit is in, all tests passing on GitHub.
Token rebrand is live on @dexscreener with the new GitHub link and a dedicated website just for OpenSAM.
One-liner install is here: rent a vps or use your own device, run one command, do a quick interactive setup, and you’ve got a 24/7 agent with persistent memory.
OpenSAM = $SAM
https://t.co/7xPsSTbAXS
for non-techys: basically, opensam is a lean alternative to openclaw
•almost the same capabilities
•~99% lighter
•runs on anything (raspberry pi, cheapest vps)
•costs few dollars a month with openrouter models
•telegram-native by default
•can even plug into @moltbook or any other @openclaw skill if you want
less overhead, same idea, way more practical
be back in few hours, feel free to test
for non-techys: basically, opensam is a lean alternative to openclaw
•almost the same capabilities
•~99% lighter
•runs on anything (raspberry pi, cheapest vps)
•costs few dollars a month with openrouter models
•telegram-native by default
•can even plug into @moltbook or any other @openclaw skill if you want
less overhead, same idea, way more practical
be back in few hours, feel free to test
Quick before I crash out to sleep.
@openclaw is cool. Moltbook is a fun experiment.
But let’s be honest, “AGI” right now is AI slop with human prompting influence.
So yeah, my "fuck you" was building a ~5k LOC Rust agent. ~99% lighter. Telegram-native. OpenRouter-ready.
Same idea. Fewer illusions.
No new token. No hype.
OpenSAM. For $SAM.
https://t.co/KhSc6LK2S9
Good night.
Quick before I crash out to sleep.
@openclaw is cool. Moltbook is a fun experiment.
But let’s be honest, “AGI” right now is AI slop with human prompting influence.
So yeah, my "fuck you" was building a ~5k LOC Rust agent. ~99% lighter. Telegram-native. OpenRouter-ready.
Same idea. Fewer illusions.
No new token. No hype.
OpenSAM. For $SAM.
https://t.co/KhSc6LK2S9
Good night.
Quick before I crash out to sleep.
@openclaw is cool. Moltbook is a fun experiment.
But let’s be honest, “AGI” right now is AI slop with human prompting influence.
So yeah, my "fuck you" was building a ~5k LOC Rust agent. ~99% lighter. Telegram-native. OpenRouter-ready.
Same idea. Fewer illusions.
No new token. No hype.
OpenSAM. For $SAM.
https://t.co/KhSc6LK2S9
Good night.
I’ve been quiet for a few days, and that wasn’t accidental.
Sometimes you need to step back and observe instead of forcing momentum.
Doing that made things pretty clear.
The space moved fast, faster than expected. Tools like clawdbot (@openclaw) didn’t just iterate, they compressed months of roadmap into days and captured gravity immediately. That doesn’t mean SAM or Codec were wrong ideas. The direction was right. But timing decides what survives, and timing won here.
Codec never really reached escape velocity, and continuing now would be more about attachment than conviction. In startup terms, that’s how you burn time instead of building leverage.
One thing I want to be very clear about: this isn’t a rug, and it couldn’t be one even if I wanted it to be. I didn’t have a bag to dump. Everything I had came from creator rewards, and a lot of that went straight back into infra, API costs, hosting, and buybacks. Anyone who followed closely knows the top PnL wallets rotated out long ago.
This was building, not farming.
This isn’t drama or a meltdown either. It’s a clean stop. Sometimes the smartest move is knowing when not to push uphill just because you’ve already walked far.
I know some of you got attached, I did too. You don’t build for months without caring. But real building also means closing chapters before they decay, taking the lessons, and moving forward sharper.
I’m not disappearing. Just resetting.
Next thing comes from a better angle, with better timing.
Appreciate everyone who was here early.
On to the next build.