Andy Tseng4m ago
Always great to hear @[email protected]'s insights. This is going to be fascinating. #LMBSeminar #MRCLMB #Science #Biology #WhiteheadInstitute #MIT #ResearchSky

RE: https://bsky.app/profile/did:plc:fod5s5sjpftlmk4e2euu5hiq/post/3mhid7e323s2n
Show threadknizer1d ago
#BostonWeekend Saturday - Balinese music - Double Gamalan afternoon #mebarung at #MIT https://calendar.mit.edu/event/gamelan-showdown-a-balinese-battle-of-the-bands
Gamelan Showdown: A Balinese Battle of the Bands

Events Calendar
Eth[̲̅a]n H[̲̅a]nsen 1d ago

Based on MIT Police scanner traffic there may be an ICE presence on or near MIT's campus in the area of 77 Massachusetts Ave in Cambridge

MIT PD traffic is not encrypted and is carried on the CoMIRS Cambridge digital trunking system for those wondering. Working on pulling recording.

#Cambridge #Cambridgema #MIT #ICE

Curated Hacker News2d ago

Parallel Perl – Autoparallelizing interpreter with JIT

https://perl.petamem.com/gpw2026/perl-mit-ai-gpw2026.html#/4/1/1

#ai #mit

Perl mit AI — Richard Jelinek — GPW 2026

Pirates.BZ Tech Startup News2d ago
MIT researchers have developed a new method to identify when large language models are confidently wrong. By comparing a model's responses to those from similar AI systems, their approach better detects unreliable predictions than traditional self-consistency checks. The technique could improve reliability in high-stakes fields like healthcare and finance. https://news.mit.edu/2026/better-method-identifying-overconfident-large-language-models-0319 #Tech #Startup #News #AI #MIT
A better method for identifying overconfident large language models

A new technique for assessing the reliability of a large language model’s predictions is better at identifying when the model is confident, but wrong. This more accurate method for uncertainty quantification could help users know whether to trust the model’s outputs.

MIT News | Massachusetts Institute of Technology
Wilma_Wein2d ago

Warum ich keine Red Hat, SUSE oder Ubuntu und deren Derivate nutzen werde wird in diesem Artikel klar. https://www.heise.de/news/Think-of-the-children-Linux-Community-gegen-die-Altersverifikation-11211944.html

#Linux #Suse #RedHat #ubuntu #mit #kubuntu #lubuntu #mint

Linux soll Ausweise prüfen: Wie die Community sich dagegen wehrt

Mehrere US-Bundesstaaten erzwingen künftig eine Altersverifikation für Nutzer. Wie sollen und können Linux-Distributionen darauf reagieren?

heise online
Pirates.BZ Tech Startup News2d ago
MIT researchers have developed generative AI models that enable robots to see through walls using Wi-Fi signals. The system reconstructs hidden 3D objects by analysing reflected wireless signals and filling in missing shape data. This could help warehouse robots verify packed items and improve smart home systems. https://news.mit.edu/2026/generative-ai-improves-wireless-vision-system-sees-through-obstructions-0319 #Tech #Startup #News #AI #MIT
Generative AI improves a wireless vision system that sees through obstructions

Wave-Former is a new system that can complete the shape of a hidden 3D object or reconstruct the scene of an entire interior room using reflected wireless signals.

MIT News | Massachusetts Institute of Technology
AIagent.at 🤖 AI News2d ago
MIT researchers are using generative AI to help robots see through walls. A new system uses Wi-Fi signals to reconstruct hidden 3D objects with 20% greater accuracy than previous methods. The technique could improve warehouse robots and smart home devices. https://news.mit.edu/2026/generative-ai-improves-wireless-vision-system-sees-through-obstructions-0319 #AIagent #AI #AIResearch #MIT
Generative AI improves a wireless vision system that sees through obstructions

Wave-Former is a new system that can complete the shape of a hidden 3D object or reconstruct the scene of an entire interior room using reflected wireless signals.

MIT News | Massachusetts Institute of Technology
AIagent.at 🤖 AI News3d ago
MIT created a method to detect when AI is confidently wrong. Comparing answers across different LLMs flags hallucinations better than self-checks. https://news.mit.edu/2026/better-method-identifying-overconfident-large-language-models-0319 #AIagent #AI #GenAI #AIethics #MIT
A better method for identifying overconfident large language models

A new technique for assessing the reliability of a large language model’s predictions is better at identifying when the model is confident, but wrong. This more accurate method for uncertainty quantification could help users know whether to trust the model’s outputs.

MIT News | Massachusetts Institute of Technology
John McLear4d ago

ayyyyyyyyy yo! I'm finally done w/ RF tests & developing #ESPhome firmware for different #NFC modules.. That's taken me ~120 hours..

All has gone to #FOSS under #MIT license...

Happy hacking people :) ❤️