OdysseyBench tests true agentic competence: • Contextual recall (across ~5K tokens)
• Multi-step reasoning
• Cross-app state tracking
• Intent inference through casual, noisy dialogue
• Tool execution (simulated, but grounded) It’s not toy data. It’s full workflow
RESEARCH
-
OdysseyBench Evaluates Core AI Agentic Competencies
By
–
-

OdysseyBench and HOMERAGENTS: A Multi-Agent Generation Pipeline
By
–
OdysseyBench is built using HOMERAGENTS, a multi-agent generation pipeline: • HOMERAGENTS+: Converts atomic tasks into multi-turn workflows • HOMERAGENTS-NEO: Synthesizes entirely new tasks using simulated app exploration Each task is: • Goal-driven
• Dialogue-based
• -

OdysseyBench: A New Benchmark for Long-Horizon AI Agent Tasks
By
–
What is OdysseyBench? A long-horizon benchmark of 602 multi-day, multi-application tasks: • Word, Excel, PDF, Email, Calendar
• 300 tasks from OfficeBench (restructured)
• 302 brand-new, complex tasks Each task is wrapped in realistic user–assistant dialogue over 5 days. -
OdysseyBench Introduced to Evaluate Long-Running AI Agent Capabilities
By
–
Today’s agents are “good” at: • Summarizing one doc
• Responding to one email
• Updating one file But in the real world? Tasks last 5 days, jump between apps, involve partial context, and require real memory. That’s what OdysseyBench simulates. -

OdysseyBench evaluates the real-world performance of LLM agents
By
–
This benchmark might kill your belief that LLM agents are “almost there.” OdysseyBench doesn’t test if an agent can summarize a file. It tests if it can survive a week at your office. And it exposes a brutal truth about today’s agents. Here’s what you need to know:
-

Luc Julia Accused of Lying About Creating Siri
By
–
LUC JULIA IS A LIAR. For years, he has been passing himself off as the creator of Siri. Tom Gruber, Siri co-founder, just confirmed that he never participated in it.
-
AI tools in clinical practice may erode doctors’ core skills
By
–
A new study in The Lancet suggests that frequent reliance on #artificialintelligence tools in clinical practice may erode core clinical skills among doctors, potentially jeopardising patient outcomes. https://
bit.ly/3UBfUmm
#GenAI #Healthcare #AI #TheLancet -
Brain chip decodes inner monologues in real time
By
–
Groundbreaking brain chip decodes people’s inner monologues in real time https://
independent.co.uk/tech/brain-chi
p-reads-inner-thoughts-speech-b2809479.html
…. #technology #Innovation #MentalWellness #technews2000 -

Day 2 at Deep Learning Indaba in Kigali: Inspiration and African AI Community
By
–

Day 2 at the @DeepIndaba in Kigali was full of inspiration and energy! 🌍 Over 1,000 African researchers, developers, and engineers are here to learn, debate, and share. #DLI2025
→ View original post on X — @shakir_za, 2025-08-19 07:09 UTC
-

AI Field Becoming Obsolete: PhD Students Should Choose Specialized Niches
By
–
""AI itself is going to be gone by the time you finish a Ph.D. Even things like applying AI to robotics will be solved by then. So either get into something niche like AI for biology, which is still in its very early stages, or just don't get into anything at all," Tarifi said."