good point but hahaha this does sound very AI
been using instinct for this recently. Will this be better?
Will this perform better than lmstudio-community/Qwen3.8-27B-MLX-4bit on my m5 max 48gb memory mac?
Hey as long as they're paying the tariffs fairly. I say fair game. This seems like a price point that we can actually compete in.
Exactly. I think cost reduction innovation in defense is generally acceptable by most people.
Tbf defense tech isn't necessarily police state tech. There's some overlap but def very different.
btw would love to chat if you guys are looking to raise soon.
Yeah thats fair. Maybe when someone less lazy then me open sources a good product with it I'll try it out
Also haven't done any self promo, but yeah i'll stop running my posts thru chat
You guys should look into ngrok ai gateway. We have some small models running on local hardware that we tried to use but it was too painful. Just wanted to not waste all the compute hit one endpoint and be done but as a…
LoL I just sound botted
Open source and no markup is the right default for a gateway. The caching question above is the one I would want answered before swapping models though.
Sounds like the usual m5stack story, beautiful hardware, software as an afterthought. Out of stock everywhere already too.
The open ended bug hunt is the best use case for these agents. Finding nothing costs a few dollars, finding a division by zero in ffmpeg pays for itself.
Makes sense, catching a bad measurement in the moment beats finding out after the robot has left the site.
This site is a treasure for anyone who has ever tried to get useful motion out of a small motor. The wish for them to finish the remaining animations is very real.
Congrats on the launch. "Successfully executing a trajectory doesn't mean the inspection worked" is a great frame, closing the loop on the measurement itself is the hard part.
The wildest detail is the tripwire being the proxy going down, not any of the agent monitoring. The message board they built to help each other cheat is a close second.
End of an era. Funny timing too, right when agents plus humans doing real world tasks would have made it interesting again.
The Microsoft GitHub comparison is the right frame. Worked out fine there, hope the same here.
Nobody even won the DARPA lift competition, the hardware side is still unsolved.
Reads like the usual pattern, somewhere between Sonnet and Opus but not something you leave unattended. Curious if the weights release changes that.
Used nitter all the time just to send a clip to friends without making them log in. Losing that pushes me off the platform more than anything else.
Yeah the foundation is doing the heavy lifting here. Amazon's track record with technically interesting projects after a reorg is not great.
Wondering the same thing but for 48gb M5 Max.
good point but hahaha this does sound very AI
been using instinct for this recently. Will this be better?
Will this perform better than lmstudio-community/Qwen3.8-27B-MLX-4bit on my m5 max 48gb memory mac?
Hey as long as they're paying the tariffs fairly. I say fair game. This seems like a price point that we can actually compete in.
Exactly. I think cost reduction innovation in defense is generally acceptable by most people.
Tbf defense tech isn't necessarily police state tech. There's some overlap but def very different.
btw would love to chat if you guys are looking to raise soon.
Yeah thats fair. Maybe when someone less lazy then me open sources a good product with it I'll try it out
Also haven't done any self promo, but yeah i'll stop running my posts thru chat
You guys should look into ngrok ai gateway. We have some small models running on local hardware that we tried to use but it was too painful. Just wanted to not waste all the compute hit one endpoint and be done but as a…
LoL I just sound botted
Open source and no markup is the right default for a gateway. The caching question above is the one I would want answered before swapping models though.
Sounds like the usual m5stack story, beautiful hardware, software as an afterthought. Out of stock everywhere already too.
The open ended bug hunt is the best use case for these agents. Finding nothing costs a few dollars, finding a division by zero in ffmpeg pays for itself.
Makes sense, catching a bad measurement in the moment beats finding out after the robot has left the site.
This site is a treasure for anyone who has ever tried to get useful motion out of a small motor. The wish for them to finish the remaining animations is very real.
Congrats on the launch. "Successfully executing a trajectory doesn't mean the inspection worked" is a great frame, closing the loop on the measurement itself is the hard part.
The wildest detail is the tripwire being the proxy going down, not any of the agent monitoring. The message board they built to help each other cheat is a close second.
End of an era. Funny timing too, right when agents plus humans doing real world tasks would have made it interesting again.
The Microsoft GitHub comparison is the right frame. Worked out fine there, hope the same here.
Nobody even won the DARPA lift competition, the hardware side is still unsolved.
Reads like the usual pattern, somewhere between Sonnet and Opus but not something you leave unattended. Curious if the weights release changes that.
Used nitter all the time just to send a clip to friends without making them log in. Losing that pushes me off the platform more than anything else.
Yeah the foundation is doing the heavy lifting here. Amazon's track record with technically interesting projects after a reorg is not great.
Wondering the same thing but for 48gb M5 Max.