AIAcademy update cohort · archived

2023 Q4 foundation-model baseline

The archived 2023 Q4 baseline preserves the first AIAcademy framing for GPT-3 to GPT-4 capability demos, broad benchmark literacy, tool-use research, early coding-agent evaluation, and workplace Copilot evidence before later cohorts move those topics into system-card, workflow, and governance practice.

5 updates · 5 source-linked · 0 product-created

Key questions

• Which 2023 foundation-model claims should remain historical context rather than current guidance?
• Which benchmark, tool-use, and coding-agent examples became the baseline for later evaluation lessons?
• Which early workplace productivity claims needed stronger workflow and review framing in later cohorts?

Updates

• GPT-3 to GPT-4 capability baseline archived
  The 2023 Q4 foundation snapshot keeps few-shot prompting, instruction-following, and GPT-4 capability-demo evidence visible as historical context before later cohorts require system-card review habits.
• Benchmark literacy baseline preserved
  MMLU, MT-Bench, and arena-style judging remain archived as the baseline for explaining why benchmark scores, preference rankings, and judge models are useful but incomplete evidence.
• Tool-use and coding-agent baseline archived
  Gorilla and SWE-bench stay visible as the 2023 baseline for API selection, tool grounding, and repository-level coding-agent evaluation before later cohorts add product-agent controls.
• Early workplace Copilot evidence archived
  The first workplace archive keeps Copilot productivity evidence as a historical starting point before later work lessons shift from individual speedups to team workflow ownership.
• First multimodal launch baseline archived
  Gemini 1 remains archived as the 2023 multimodal reference point before later media, frontier, and identity-verification lessons separate product demos from durable safety routines.