> Students in the Codex group made more progress during the seven sessions of training and finished significantly more tasks (91%) than the Baseline group (79%).
I expected this
> On the Immediate Post-test that was conducted a day after the training sessions, both groups performed similarly on coding tasks (Codex: 61%, Baseline: 63%) and multiple-choice questions (Codex: 49%, Baseline: 42%).
> However, on the Retention Post-test that was conducted a week later, the Codex group was able to retain what they learned slightly better on the coding tasks (Codex: 59%, Baseline: 50%) and multiple-choice questions (Codex: 44%, Baseline: 35%).
However, this completely surprised me. I assumed there would be an over-reliance on AI in the test group
I find the analysis of the Codex group much more interesting and I feel further studies are necessary before drawing conclusions.
Students frequently (n=501, 30%) copied the task description to generate the entire code with no prior manual coding attempts.
Sometimes (n=197, 12%) students divided the task into multiple subgoals, and asked the AI to generate *only* the first subgoal instead of the entire task.
When decomposing the task into multiple subgoals, students sometimes (n=85, 5%) asked for code that was already in their editor.
Although rarely (n=16, 1%), but sometimes students generated code after having the solution to check and compare the AI's output with their own solution.
Students occasionally (n=89, 5%) wrote prompts that were similar to pseudo-code (e.g. "for num in numbers, if num > large, set large to num").
Although most of the times students properly tested the AI-generated code before submitting, there were several (n=63, 4%) instances in which students submitted AI code without testing it.
Although rarely, but sometimes (n=30, 2%) students actively tinkered with the AI-generated code to properly understand the syntax and logic.
Similarly, sometimes students manually added code (like `print` statements) to the AI-generated code to help them verify that it works correctly.
Anecdotal evidence based on my own experience has suggested better recall of things I've engaged with and learned through ChatGPT than with standard learning through books or videos. I think the interactivity is an important aspect.
The AI seems to be more interested in winning than answering questions - it doesn't really challenge ideas and seems to just make statements that are related to what was mentioned before, but this was fun to play with
I've done all the static site generators, I just don't post enough to justify the maintenance, if you do they're probably a better solution! As is, it's pretty basic but I want to add features that GitHub pages don't support
Creating a GitHub repo with one file titled “index.md” and publishing it to GitHub Pages creates a single-page site using relatively sane defaults. There's no maintenance involved whatsoever. You can add multiple pages as well by creating new .md files and publishing them.
No, GitHub pages are tied to "an" account. It can be a secondary "anonymous" account. You can create an anonymous bear blog or one of the other million free writing and zero maintenance tools. It's certainly more anonymous than having to disclose your real identity via a credit card on a random person's page.
Also, what sort of content do you hope that your target audience (Bob the landscaper who is non technical but has a super urgent desire to send a message to the world in an anonymous way) will publish?
Yeah but would I tell my mom to pay a dollar on this site? Would she trust it? Does she need a website or does an email/google doc satisfy her needs. What are the use cases for entirely anonymous sites that you have in mind?
Last time I tried, there is no verification other than and email link for Github accounts. I think you could make an account and publish with just a throwaway email.
Yeah, honestly there are similar things and other things that do this better if your goal is to just run a blog. This was a simple project I figured I could finish in two weeks and refresh some skills along the way while using some new libraries and technologies.
I expected this
> On the Immediate Post-test that was conducted a day after the training sessions, both groups performed similarly on coding tasks (Codex: 61%, Baseline: 63%) and multiple-choice questions (Codex: 49%, Baseline: 42%).
> However, on the Retention Post-test that was conducted a week later, the Codex group was able to retain what they learned slightly better on the coding tasks (Codex: 59%, Baseline: 50%) and multiple-choice questions (Codex: 44%, Baseline: 35%).
However, this completely surprised me. I assumed there would be an over-reliance on AI in the test group