Recent Posts

• The programmer's climb
  by gandalfgeekon 4/27/2025, 4:24:34 PM with 0 comments
• Dijkstra would've (incorrectly) HATED vibe-coding
  by gandalfgeekon 4/12/2025, 7:15:38 PM with 1 comment
• "But LLMs are not deterministic "
  by gandalfgeekon 3/29/2025, 10:47:32 PM with 3 comments
• The case against local LLMs
  by gandalfgeekon 3/29/2025, 6:24:58 PM with 0 comments
• Human evals are outsourced vibe checks
  by gandalfgeekon 2/23/2025, 5:40:53 PM with 0 comments
• Spec-Driven Vibe-Coding
  by gandalfgeekon 2/9/2025, 5:57:06 PM with 1 comment
• AI resistance considered harmful (for programmers)
  by gandalfgeekon 1/17/2025, 8:55:07 PM with 3 comments
• I use AI (late 2024)
  by gandalfgeekon 11/21/2024, 4:38:54 AM with 1 comment
• Subject-object-verb, or– how we use AI
  by gandalfgeekon 7/1/2024, 2:24:03 AM with 0 comments
• The De-Nerdification of Programming
  by gandalfgeekon 3/30/2024, 5:39:09 AM with 0 comments
• Prompting Is Programming
  by gandalfgeekon 9/25/2023, 4:08:35 PM with 0 comments
• Metaphors for Thinking about LLMs
  by gandalfgeekon 8/21/2023, 6:37:15 PM with 0 comments
• How do experienced programmers use Copilot?
  by gandalfgeekon 7/24/2023, 4:46:41 PM with 0 comments
• Do large context windows for LLMs help?
  by gandalfgeekon 7/10/2023, 4:57:22 PM with 0 comments
• How do novice programmers use Copilot?
  by gandalfgeekon 7/5/2023, 5:19:39 PM with 0 comments
• Faster Python with Static Types
  by gandalfgeekon 6/26/2023, 9:42:02 PM with 0 comments
• Tree of Thought prompting for LLM reasoning
  by gandalfgeekon 6/12/2023, 6:25:16 PM with 0 comments