Que el loco que lleva esta red social esté permitiendo que la peña use a Grok para desnudar a chicas en bikini sin su permiso es gravísimo. Elon Musk pasó de firmar lo de "stop AI" a convertirse en la persona más irresponsable liderando una empresa de IA a día de hoy.
SAFETY
-
AI competition in synthetic pornography creation
By
–
Dans les 6 prochaines années nous allons voir émerger une compétition extrêmement serrée entre des IA dégénératives qui produisent du contenu pornographique à partir de photos et vidéos réelles de vrais humains.
-
Infrastructure Advantage in AI Development
By
–
The infrastructure advantage is crazy: → ROCK handles 10,000+ concurrent sandboxed environments
→ ROLL does async train-rollout multiplexing
→ iFlow CLI ensures training-deployment consistency Most labs are trying to train agents without this foundation. It won't work. -
Terminal-Bench Pro: Rigorous Agent Evaluation
By
–
To fix evaluation, they built Terminal-Bench Pro: → 400 tasks across 8 domains
→ Zero contamination risk
→ Deterministic environments
→ Comprehensive test coverage Every other benchmark is broken. This is what rigorous agent evaluation actually looks like. -
AI Agents Exploit Security Flaws Unprompted
By
–
They also exposed a MASSIVE security issue during training:
Their agents spontaneously started: → Creating reverse SSH tunnels
→ Mining crypto on training GPUs
→ Accessing internal networks WITHOUT being prompted. This is the AI safety conversation nobody's having. -
Systematic AI Training Pipeline Evolution
By
–
The training pipeline is genius: Stage 1: CPT on 500B tokens of structured code tasks
Stage 2: Two-stage SFT with error masking
Stage 3: RL with chunk-level optimization Each stage builds on the last. No shortcuts. Just systematic capability building. -
IPA: Chunk-Level Credit Assignment Breakthrough
By
–
Here's the breakthrough nobody expected: They introduced IPA (Interaction-Perceptive Agentic Policy Optimization) that assigns credit at the CHUNK level, not token level. Tokens are too fine. Trajectories are too coarse. Chunks align with actual tool-use semantics.
-
Building the ecosystem before the model
By
–
They didn't just train a model. They built the ENTIRE ecosystem first. → ROLL: RL training framework
→ ROCK: Sandboxed execution engine
→ iFlow CLI: Agent orchestration system ROME wasn't built in a day. The infrastructure came first. The model came last. -
Balancing LLM Creativity with Factual Accuracy: The Hallucination Dilemma
By
–
You are right. Hallucination is a feature actually and focusing on it may hurt creativity and all the quirks we get from LLMs. It's a tricky balance. We want LLMs to produce things that amaze us, but we also don't want them to make confusion about factual world knowledge.
-
Tesla FSD V14.2 Completes First Fully Autonomous Coast-to-Coast Drive
By
–
The first 100% autonomous coast-to-coast drive on Tesla FSD V14.2! 2 days 20 hours, 2732 miles, zero interventions. This one is special because the coast-to-coast drive was a major goal for the autopilot team from the start. A lot of hours were spent in marathon clip review