Zhemal Khamidun
Head of AI at Alpina Digital, CPO of AlpinaGPT (~30-person engineering & AI team). Founder of Vibe Academy & Hamidun Consulting, author of the AI channel «Готовим ИИшницу» (@jemal_hamidun).
OpenAI Hacked by an AI Agent: Hugging Face CEO Demands Radical Transparency
Hugging Face CEO Clément Delangue called the OpenAI breach the first cyberattack by an autonomous AI agent and urged the industry to embrace radical transparency.
Relay-Bench: Even GPT-5.5 Scores Just 43.3% on a Multi-Domain Reasoning Test
The new Relay-Bench benchmark tests how AI handles linked tasks from different fields in a single prompt. Even the strongest model, GPT-5.5, solves fewer than half…
SIFT: A Self-Improving Document Classifier That Retrains Without a Human
SIFT is a document classification service in which an expensive LLM teaches a cheap CPU classifier. The corpus grows from live traffic, and two safety gates allow…
PEARL: A 4B Model Beats DeepSeek-V3.2 685B at Optimization Modeling
SmolLM2-1.7B Beats 34B Models at MLIR Code Generation Without Fine-Tuning
Phionyx: A Deterministic AI Runtime That Treats LLM Outputs as 'Sensor Readings,' Not Decisions
Latency-Aware LLM Query Routing: Up to 40% Better Accuracy-Cost Utility at the Same Latency
SAAG: A Three-Stage Method for Diagnosing Function-Calling Errors in AI Agents
SAAG splits the evaluation of AI agents' function calls into three stages and shows where exactly the agent went wrong, instead of returning a single right/wrong…
MILP-Evo: An LLM Designs and Speeds Up the Logic of MILP Optimization Solvers
In a closed loop, a large language model designs and improves the logic of MILP solvers on its own, producing readable code instead of a black box, according to the…
How to Measure Residual Risk in Autonomous AI Agents: The CPSAINT and FRIESA-K Framework
BatchDAG: LLM Planner Builds an Operation Graph and Cuts Model Calls by Up to 47x
Thinky Releases Open Inkling 975B-A41B Model Under Apache 2.0 License
Evidence Chain Evaluation: An AI Fact-Checker That Is Allowed to Say "Not Sure" Instead of Issuing False Verdicts
Xaira unveils X-Cell, a 4.9B-parameter "virtual cell" for drug discovery
Xaira trained X-Cell, a 4.9B-parameter "virtual cell", on the largest genetic perturbation dataset yet, betting that causal models need causal data.
Linus Torvalds on AI in Linux: a useful tool, and dissenters can fork
Linux creator Linus Torvalds says the kernel will not be an anti-AI project: AI is a useful tool, and anyone who disagrees can fork the project or leave.