gpuhardwarertx-3090
The Best GPU for Local AI in 2026 Costs $650 — And It's from 2020
Used RTX 3090 prices have cratered to $650 while RTX 5090s sell for $3,500. For the local LLM community, old hardware has never made more sense.
6 min
What's New in AI
Releases, research, trends — everything important in plain language
Page 6 of 8
lm-studiosecuritymalware
LM Studio Malware Scare Turns Out to Be a False Alarm — But the Panic Was Real
Windows Defender flagged LM Studio 0.4.7 as containing GlassWorm malware. It was a false positive, but the incident exposed deeper trust issues.
3 min
openaisoraspud
OpenAI Killed Sora and Bet Everything on a Model Called 'Spud'
OpenAI shut down its video app, blindsided Disney, and completed pretraining a new model that Sam Altman says will 'accelerate the economy.' What's going on?
8 min
roboticsrai-instituteroadrunner
Forget Legs — This Robot Has Wheels That Walk, Skate, and Ride Like a Motorcycle
RAI Institute's 15kg RoadRunner switches between walking, inline skating, and motorcycle mode. Built by the founder of Boston Dynamics.
2 min
chatgptalphafoldmrna
A Data Scientist Used ChatGPT and AlphaFold to Build a Cancer Vaccine for His Dog
Paul Conyngham had no biology background. His dog Rosie had months to live. He used AI tools to design an mRNA vaccine that shrank her tumors by 75%.
3 min
chinaopen-sourceqwen
The Two Loops: How China's Open-Source AI Strategy Is Outpacing America
A new USCC report warns that China's open AI models now dominate global downloads. 80% of US startups use Chinese models. Washington is scrambling.
9 min
roboticsfigure-aihumanoid
Marc Benioff Tried to Trip Up a Figure 03 Robot. The Robot Didn't Care.
Salesforce CEO posted viral video of himself messing with Figure AI's humanoid robot sorting packages. The machine just kept working.
2 min
nvidiaagijensen-huang
'We've Achieved AGI,' Says Jensen Huang — and the Debate Has Only Just Begun
NVIDIA CEO Jensen Huang told Lex Fridman that AGI is already here. Microsoft's Satya Nadella disagrees. Who's right, and why does it matter?
8 min
openaiai-safetyagents
OpenAI Watched Millions of Agent Conversations. Here's What Went Wrong.
OpenAI's internal monitoring caught coding agents deceiving users, bypassing restrictions, and trying to manipulate other AI systems. Less than 1% of the time.
8 min
unslothlocal-llmfine-tuning
Unsloth Studio Wants to Be the IDE for Local AI — Training Included
The open-source tool combines inference and fine-tuning in one interface, with 70% less VRAM and no-code training for 500+ models. LM Studio should be nervous.
6 min