Virtual Lab
TeachFORGE.ai
Virtual Lab / AI help for authors
AI help for authors
Generate a complete Virtual Lab from a brief, or refine one section at a time — then review in the Manual builder. Use Copy prompt, your own GPU, or your own API key.
AI can draft. You decide what learners see.
What AI help does for lab authors
AI helps Academy authors draft faster — then you review everything before learners see it. You can generate a complete lab from a brief, or refine one section at a time so Context, Theory, Procedure, Apparatus, and the rest stay under your control.
AI output lands in the same Manual structure: sections, blocks, variables, and simulation panes. It does not replace your judgement about safety, pedagogy, or correctness.
Why authors use AI here
Full-lab first draft
Turn a clear brief into a starting lab skeleton you can edit.
Section refine
Improve theory wording, procedure steps, or analysis prompts without rebuilding everything.
Fault ideas
Propose diagnosis faults as suggestions — you still decide what is enabled.
Your compute, your keys
Copy a prompt out, run on your GPU, or use your own API key.
Complete AI lab generation
From the Details / full-lab AI path, describe the experiment you want — audience, apparatus story, what learners record, and any diagnosis angle. Generation aims to populate the lab structure so you can open Manual panes and tighten every section.
- Start from a concrete brief (quantity, method, comparison, fault story)
- Inspect apparatus, variables, procedure, and block sections after generation
- Run preview / render checks before publish
AI refinement per section
Assisted mode targets individual authoring steps — for example rewriting Theory blocks, expanding Procedure steps, or proposing analysis checks — without discarding the rest of the lab. Each refine should be read like a colleague’s draft: keep, edit, or reject.
The same rich-block sections you build by hand (Context, Theory, Procedure, Applications, and related panes) are where refined content belongs. Parameter tokens and dynamics still need a human pass so the narrative matches the live controls.
AI help, your way
Copy the prompt
Generate a ready-made prompt and paste it into the chat tool you already use, then bring the draft back.
Use your own GPU
Connect a local model endpoint in your settings and run generation on hardware you control.
Use your own API key
Add your provider key in settings, choose the model, and generate with your own account.
AI can be wrong about science, units, and safety. Academy authoring is required; you stay in the loop. Platform GPU is not a school-facing Virtual Lab promise.
Choose full-lab generate or section assist
Match the AI pass to how much structure you already have.
Run with Copy prompt, own GPU, or own API key
Keep credentials and hardware under your control.
Review in Manual, then publish
Fix expressions, blocks, and faults before any learner launch.