About this Event
Is GPT-6 Astra finally reliable enough for real work? We'll dig into what holds up, what still breaks, and how to use it anyway.
AI agents are everywhere. They can browse, write, automate, even use a computer. But can you trust them with real work?
If you haven’t been tracking our events, we’ve been moving fast and switching the spotlight each month:
March: OpenClaw
April: ChatGPT Voice
May: Claude
June: Hermes
July: Obsidian
Aug: Grok Bot
Now GPT-6 Astra shifts the conversation. It’s no longer only “can AI take actions?” It’s whether the brain behind the agent is finally strong enough for messy, multi-step work with less hand-holding.
At this Fri AI Lab we’ll compare ChatGPT Work + GPT-6 Astra with Claude Cowork, Grok Bot, Hermes, and OpenClaw. We’ll look at intelligence, autonomy, memory, computer use, customisation, persistence, and quality of output.
Simple room question: Which AI agent would you actually trust with real work?
Come test, compare, and decide for yourself.
📅 Date: Sept 18 (Fri)
⏰ Time: 4–6pm PDT
📍 Venue: AIC (Alexander Innovation Center), Alexander College, 570 Dunsmuir Street, Ground Floor Presentation Room
👥 Attendance Limit: 20 people max.
🛠️ What We’ll Be Demoing
🛠️🤖 AI Agent Compare Sandbox: Build, Don’t Just Watch
This is NOT a class. It is a weekly workshop where builders share progress, troubleshoot in real time, and exchange raw ideas.
- Be a Builder: Action over observation. Bring something you want to create, improve, or compare.
- Progress Over Perfection: Your project does not need to be polished. Broken experiments and unfinished ideas are welcome.
- Practical and Verifiable: We focus on AI that solves a real problem, saves time, or helps you choose a clearer stack.
💻 What to Bring
Bring your laptop, one real task you’d hand an agent (inbox, research, scheduling, or follow-ups), and your curiosity. You do not need a technical background. Anything involving AI agents, large language models, automation, or vibe coding is welcome.
🗓️ Agenda
4:00 PM | Rapid-Fire Intros — Who are you, what are you building, and which agent would you trust with real work today? One minute each!
4:20 PM | Demo and Compare — Astra + scorecard vs Claude Cowork, Grok Bot, Hermes, OpenClaw
5:00 PM | Sandbox — run your real task on one stack; note where it needs a human boss
5:30 PM | Open Q&A and Troubleshooting — No gatekeeping. Ask anything, test ideas, get past roadblocks.
🚀 Who This Is For:
- Non-technical founders choosing a stack without buying the hype
- Small and medium-sized businesses comparing agents for real work
- Creators and solo founders who want a clear “trust with work” decision
- Curious beginners: bring a real task and leave with clarity and a next step
✅ What GPT-6 Astra is, and what “reliable enough for real work” does and does not mean
✅ Side-by-side: ChatGPT Work + Astra vs Claude Cowork vs Grok Bot vs Hermes vs OpenClaw
✅ Scorecard lenses: intelligence, autonomy, memory, computer use, customisation, persistence, output quality
✅ Honest limits: token cost, long runs that stall, access quirks, what still needs a human boss
✅ Live compare: bring one real task and pressure-test which stack you’d actually trust with real work
⏳ Update Status
If you can’t make it, please cancel 24 hours in advance. No-shows will be blacklisted from future sessions to keep things fair for those on the waitlist.
Disclaimer: By attending, participants agree to follow Alexander College facility rules, assume any risks related to the event, and release Alexander College, the event organiser, and their representatives from claims or liability arising from their attendance or use of the event room.
Note: We meet every Friday, 4–6 PM @ Alexander College. Bring a friend!
#AIAgents #LLM #BuildInPublic #AIStartup #SoloFounder #AgenticAI #AIC #AlexanderInnovationCenter #GPT6Astra #FriAILab
Event venue & nearby stays
Alexander College, 570 Dunsmuir Street, Vancouver, Canada
Concerts, fests, parties, meetups — all the happenings, one place.