Giles
gdeejay.bsky.social
Giles
@gdeejay.bsky.social
Applied economist working in international development policy
🇹🇴🇸🇧🇻🇺🇫🇯🇦🇺🇲🇲 (etc),

Mostly memes.

Sometimes post about #publicpolicy, research, data science, #PFM and #Rstats

Economist | Independent Consultant

All views my own
Pinned
Dear #rstats,

Please find attached the alpha version of my book "An Introduction to R for Policy Analysis":
policyanalysislab.com/shared/r-for...

Unfortunately, the draft has been released too late to be considered for a Nobel prize, but I look forward to being nominated in 2026.
...
an_intro_to_r_for_policy_analysis
policyanalysislab.com
I decided to drop How to Defend Australia (H2DA) and move onto something else.

I chose 'The Picture of Dorian Gray's by Oscar Wild.

I enjoyed H2DA, but I wasn't interested in going into further detail.

...
December 27, 2025 at 1:20 AM
Beach-finity
December 26, 2025 at 9:07 PM
They weren't anybody's worst nightmare.

They were just how assessments were conducted.

🤨
Does anyone else bristle at this framing? WSJ says blue books are “torturing” students with hand cramps, and “nobody likes them.”

Listen, students have been outsourcing everything to AI and cheating their way through college. Blue books should be celebrated as a return to authentic human learning.
They Were Every Student’s Worst Nightmare. Now Blue Books Are Back.
Cheating with ChatGPT has become a huge problem for colleges. The solution is painfully old-school.
www.wsj.com
December 25, 2025 at 3:36 AM
...
December 24, 2025 at 10:19 PM
Good news #researchsky, there'll be no way to avoid #AI in 2026!
December 24, 2025 at 9:38 PM
Beached
December 22, 2025 at 7:41 AM
Beached
December 22, 2025 at 7:31 AM
I wish there was a central #AI blocklist I could have my name added to.

It would save time being added to them one by one.
December 22, 2025 at 3:36 AM
I've started reading 'How to Defend Australia' by Hugh White.

I picked this as:
1.) I know little about Australian defence polict; and
2.) It's highly rated and available for free on Kindle Unlimited.

I'm enjoying it so far. It's interesting and easy to read.
December 21, 2025 at 9:04 AM
I spent the day helping out my wife at the market in 36 degree Celsius heat.

🥵

It was horrible, but I'd say the guy walking around dressed as Santa had it worse.

😬
December 21, 2025 at 5:38 AM
I'm currently reading Adam Smith's 'The Theory of Moral Sentiments'.

This is going to be a slow read (like the Wealth of Nations) as it's extremely dense, but I'm enjoying his treatment of the causes and magnitude of empathy.

I have a couple of books like this that I read in a piecemeal fashion

🤔
December 21, 2025 at 12:11 AM
Done.

⭐⭐⭐⭐
(Out of five.)

It's a beautiful book. The only negative for me was that the latter chapters felt less focussed.

Although it's not a literary masterpiece, I found the writing style made the story feel more personal and honest (like reading the memoir of a family member).

❤️
For my next book I thought I'd go with 'The Happiest Man on Earth'.

I chose this on the basis of it being short and one of the highest rated books on my Goodreads backlog.

I also wanted to go with something that would tug on my heart strings prior to Christmas.
December 20, 2025 at 3:46 AM
This has been my observation when using local #LLM models too.

Even if the model *can* handle a large amount of text it doesn't mean you *should*.

IMO it's generally better to have it process context in a piecemeal fashion and combine it than feed it evening at once.
It didn't matter how much I massaged my prompts, if I gave it too much text it messed things up. My primary LLM at the time was Gemini CLI, mostly because I find it handles text better than Codex or Claude which try to approach problems too code focused instead of LLM focused
December 19, 2025 at 11:38 PM
For my next book I thought I'd go with 'The Happiest Man on Earth'.

I chose this on the basis of it being short and one of the highest rated books on my Goodreads backlog.

I also wanted to go with something that would tug on my heart strings prior to Christmas.
December 19, 2025 at 10:11 PM
Done.

Excellent book:
⭐⭐⭐⭐
(Out of five).

Maybe the only reason i didn't give it 5/5 is that the ending felt flat, but I think that's the point. I think I need to sit and reflect on it more.

If you're a Bladerunner fan it's still worth reading. They tell different stories.
Next book is "Do Androids Dream of Electric Sheep."

I've had this one on my list for years.
December 19, 2025 at 8:06 PM
I'm *really* enjoying this.

It's familiar territory as I've seen Bladerunner but it explains a lot of things that I'd missed in the movie.
Next book is "Do Androids Dream of Electric Sheep."

I've had this one on my list for years.
December 18, 2025 at 11:38 PM
Revenge of the beach.
December 18, 2025 at 8:03 PM
If you're wondering why I'm tearing through books at the moment: I'm on holidays and have become repulsed by the computer.

😂
December 18, 2025 at 1:45 AM
Next book is "Do Androids Dream of Electric Sheep."

I've had this one on my list for years.
December 17, 2025 at 9:53 PM
Done.

This is a hard one to rate as I enjoyed it overall, but there were parts which felt a little predictable, and grating.

Maybe ⭐⭐⭐ out of five(?). I think I would have enjoyed it more as a teenager.
Claude suggested "Flowers for Algernon" as my next read, so here we go...

🤷🏼‍♂️
December 17, 2025 at 9:22 PM
Holy hell Spotify Wrapped was done so well this year!

Here are my highlights.

(As somebody that listened to Pink Floyd for most of his 20s, younger me would be ashamed.)

😬
December 17, 2025 at 8:07 AM
Yeah, I can see why people found this cringy.

I'm 1/3 the way through, but I don't think I'll finish it.
Claude suggested "Flowers for Algernon" as my next read, so here we go...

🤷🏼‍♂️
December 17, 2025 at 12:10 AM
This gave me flashbacks to the professor yelling at the class during an 'Advanced Macroeconomics' lecture because nobody knew the relationship between the interest rate and bond rate.

He was a Rhodes scholar and somehow got stuck with us morons.

I still feel bad about this.

😬
Why is the marginal cost curve U-shaped?

Student: Because it goes down and then it goes up.
December 16, 2025 at 10:00 PM
Praying that the courier delivers my coffee beans before breakfast so I don't have to drink decaf.

👼🏾🙏🏼👼🏼
December 16, 2025 at 8:14 PM
Claude suggested "Flowers for Algernon" as my next read, so here we go...

🤷🏼‍♂️
December 16, 2025 at 12:29 AM