@davidtaylor0390@davidtaylor0390 = pretends the US government is so non-violent and so righteous
10 people fatally shot by immigration officials in Trump’s 2nd term
https://t.co/bFFP0mGCSQ
Federal immigration agents have shot at least 21 people during Trump's 2nd term
https://t.co/Zq2KyfpXcn
57 juta penerima MBG melalui 27.486 SPPG. Siswa yang lapar saat sekolah dilaporkan turun dari 56% ke sekitar 16%. Angka tentang buah, protein hewani, dan rasa kenyang ada di infografik ini. #MBG
eh, I wasn’t referring to what indie hacking is about
I was saying idie hackers are about to be the first to become obsolete in the AI era
that’s one of the reasons Levels and his friends are such doomers right now, trying to convince everyone this will happen to them as well
their indie bubble is popping as we speak
💎 Say hello to LLMPad! Check out the airdrop page for $LLM, a memecoin name in the Solana crypto ecosystem where $SOL is central to the network.
CA=B5uNPxEWB6WU3Rej33iDuX7yobZ5N4kXXckoj8KSpump
https://t.co/K1PGWaIEEq
A robot navigator that was never fine-tuned for navigation just beat the ones that were.
SuperNav (Zhejiang University et al.) takes an off-the-shelf multimodal LLM, leaves its weights untouched, and wraps it in an agent harness: navigation skills, tools, task-progress tracking, and a "point at the destination in the image" interface that hands the actual motion to navigation tools.
Results in simulation:
• Single-object search: 78% success vs 34% for the best of 4 baselines
• Multi-object: 59.5% vs 37.5%
• Demand-driven (need-based) requests in AI2-THOR: 73.3% vs 62.5%
It also runs on a real Unitree Go2 quadruped.
The pattern keeps showing up: for embodied AI, the harness may matter more than the fine-tune.
https://t.co/L4FbokDfLw
#Robotics #EmbodiedAI #AIResearch
@DREAMS_137@_posthumanist_@film_girl Wah, analogi sungainya keren banget sih. Jadi LLM itu kayak aliran air yang udah dibikin jalurnya sama manusia ya, terus kita cuma ngatur dikit buat ngarahin.